Fitness of Multi-Resolution Remotely Sensed Data for Cadastral Mapping in Ekiti State, Nigeria

Abstract

In developing nations, such as Ekiti State, Nigeria, the utilization of remotely sensed data, particularly satellite and UAV imagery, remains significantly underexploited in land administration. This limits multi-resolution imagery’s potential in land governance and socio-economic development. This study examines factors influencing UAV adoption for land administration in Nigeria, mapping seven rural, peri-urban, and urban sites with orthomosaics (2.2 cm to 3.39 cm resolution). Boundaries were manually delineated, and parcel areas were calculated. Using the 0.05 m orthomosaic as a reference, the Horizontal Radial Root Mean Square Error (RMSEr) and Normalized Parcel Area Error (NPAE) were computed. Results showed a consistent increase in error with increasing resolution (0.1 m to 1 m), with RMSEr ranging from 0.053 m (formal peri-urban) to 2.572 m (informal rural settlement). Formal settlements with physical demarcations exhibited more consistent values. A comparison with GNSS data revealed that RMSEr values conformed to the American Society for Photogrammetry and Remote Sensing (ASPRS) Class II and III standards. The research demonstrates physical demarcations’ role in facilitating cadastral mapping, with formal settlements showing the highest suitability. This study recommends context-specific imagery resolution to enhance land governance. Key implications include promoting settlement typology awareness and addressing UAV regulatory challenges. NPAE values can serve as a metric for assessing imagery resolution fitness for cadastral mapping.

1. Introduction

Efficient land administration is fundamental to governance and sustainable development [1,2,3]. While land administration involves a range of processes and functions, which includes land registration, cadastral mapping, land tenure systems, land valuation, and land use planning, the primary goal of land administration is to establish and maintain accurate and up-to-date records of land ownership, rights, and usage, thereby facilitating efficient land management and supporting socio-economic development [3,4,5,6]. This process encompasses establishing and maintaining a systematic record of land parcels, defining rights and obligations of individuals and groups to the land, and recording rights, which relies significantly on the cadastral mapping approach adopted. Hence, there is a persistent need for improved methods to aid and fast-track effective and consistent measurement and update of land records in the most appropriate way.

Cadastral mapping is a systematic process of delineating, recording, and managing land parcels and property boundaries within a specific geographic area [7,8]. It involves creating maps or plans that represent the distribution of ownership, use, and related rights [9,10]. Cadastral mapping employs various methods and technologies to accurately represent property and parcel boundaries. Cadastral mapping relies significantly on the spatial framework of land administration.

With improvements in technologies, the procedure of cadastral mapping has advanced from pacing to chains, tapes, electronic distance measurement, satellite navigation systems, and aerial imagery. Two prevalent technologies for cadastral mapping are the Global Navigation Satellite System (GNSS) and aerial imagery [8,11,12,13]. While GNSS technology enables surveyors to precisely determine the location of points, aerial imagery, satellites, aircraft, and Unmanned Aerial Vehicles (UAVs) provide visual information for boundary delineation alongside information on land use, vegetation health, land cover, topography, and other environmental and geographic features. GNSS allows for efficient and rapid data collection in the field. On the other hand, aerial imagery contributes to comprehensive data capture. Furthermore, aerial imagery also facilitates remote data collection when necessary. GNSS is an integral part of aerial imagery collection as UAVs often carry GNSS chips; GNSS are used in coordinating ground control points (GCPs) for UAV imagery in obtaining control points for aerial photo mapping and georeferencing satellite imagery [14].

Orthomosaics acquired from Unmanned Aerial Vehicle (UAV) mapping processes, aerial images obtained from aircraft, and satellite imagery are the three commonly utilized aerial imagery for cadastral mapping. The Fit-for-Purpose Land Administration (FFPLA) approach advocates for the use of aerial imagery for land administration due to its cost-effectiveness and scalability [4,15]. This approach is seen as a means of accelerating the process of land registration, ensuring the security of tenure for the poor, achieving global coverage of land registration, and providing individuals and groups with the benefits of an efficient land administration system. Based on principles such as flexibility, inclusiveness, participation, affordability, reliability, upgradeability, and attainability, the FFPLA approach seeks to improve the spatial, legal, and institutional frameworks of land administration to aid effective land governance for development [4,16]. While the FFPLA does not necessarily mean inaccurate surveys, it prioritizes coverage and scalability above precision. In the FFPLA context, “Accuracy relates to the purpose rather than technical standards” [17]. Meanwhile, with the prevalence of high-resolution remotely sensed data, notably from UAVs, aerial imagery can now attain accuracies on par with those required for conventional cadastral mapping.

Despite global advancements in UAVs and satellite imagery for land administration [7,18], their underutilization persists in Nigeria and some other countries of the world where the need to fast-track land registration has become more prominent. A major challenge to leveraging remote sensing data for land administration previously was spatial accuracy because of the resolutions of available imagery. However, advancements in remote sensing technology have led to an increase in the resolution of aerial imagery, making it more accurate and reliable for land administration purposes. This should remove the previous challenge and encourage increased utilization of this technology. Unfortunately, this is not the situation, as most states in Nigeria and similar places in developing countries still do not leverage the advantage. This underutilization adversely affects land administration. In addressing the intricate dynamics surrounding the limited adoption of UAVs in Nigeria, it becomes important to highlight the potential synergy of integrating UAVs with GNSS for land administration and broader land management functions.

The availability of high-resolution aerial imagery across all contexts is another significant factor. The use of UAVs—the major source of high-resolution imagery—is one of the most regulated technologies in the world, owing to their dexterity and use for various purposes [19,20]. Except for state-implemented projects, UAV regulations in Nigeria often limit the use of UAVs for mapping rural and peri-urban areas. Obtaining permission to use UAVs for mapping in urban areas in Nigeria is a complex and time-consuming process, often making it impractical or impossible. Meanwhile, it is commonly assumed, based on building density, that urban areas require high-resolution imagery for mapping, while rural areas can use low-resolution data. Unfortunately, rural and urban contexts are not exclusively different from each other. There are urban settlements with some rural patterns in between and rural settlements with some densely populated settlements. The need to determine the fitness of the various types of remotely sensed aerial images in mapping the different contexts of informal/formal rural, peri-urban, and urban areas for land administration underpins the need for this research. Additionally, the underutilization of UAV technology in Nigeria and similar developing countries, owing to several unclear factors, and the regulatory restrictions against the use of UAVs in urban areas where high-resolution images are most required for land administration and management owing to population density buttresses the need for this research.

Several researchers have advocated for the use of UAV orthomosaics and satellite imagery for land administration. Notably, the potential of unmanned aerial system technology, including UAVs, was advocated to meet land data and user needs in Rwanda. The study concludes that UAS can contribute significantly to match most of the prioritized needs in Rwanda while acknowledging the limits posed by structural and capacity conditions [21]. Guidance on achieving efficient and reliable UAV data acquisition by analyzing various flight configurations and their influence on data quality and cadastral feature extractability was also presented by earlier research [12,22]. The work provides insights into the optimal number of ground control points required for accurate mapping and the challenges faced by different land-use categories. It recommends the use of drones with high-quality optical sensors, a suitable number of overlap ratios, and an appropriate number of ground control points for reliable cadastral mapping.

Origins and debates surrounding the use of remote sensing technologies, including UAV and satellite imagery for land administration, were also presented in earlier research that discusses how remote sensing can be an entirely legitimate, if not an essential, part of the domain [23]. The research concludes that photogrammetric and remote sensing methods have a strong historical and contemporary presence in land administration practice. Ground methods continue to dominate in many jurisdictions. The review concludes that any remnant arguments on the use and apparent limitations of photogrammetric methods and remote sensing applied to land administration can hardly be sustained. The use of UAVs has been identified as a promising tool for land administration in rural, peri-urban, and urban contexts. However, the success of UAVs in land administration is contingent upon the technology’s ability to address the specific challenges and needs of a given context. In areas where land administration may be absent or incomplete, it is important that flexible and pragmatic approaches be adopted to meet the specific needs of communities and governments.

Despite advancements in aerial imagery resolution, the integration of this technology into cadastral surveying practices in Ekiti State, Nigeria, and other developing regions remains limited [15]. This aligns with global trends, where UAV-based cadastral surveying faces implementation challenges [23,24,25,26]. Notably, Edo State, Nigeria, has successfully utilized satellite imagery, aerial orthophotos, and UAV imagery for land administration and management through the Edo GIS.

The traditional GNSS process, while precise, can be costly and time-consuming, requiring visits to every demarcation point. Aerial imagery offers a potential solution to speed up cadastral mapping processes [9,27,28,29,30]. However, UAV adoption in Ekiti State is hindered by regulatory restrictions and technical limitations [15].

Previous research highlights UAV benefits, including increased efficiency, accuracy, and cost-effectiveness. Yet, the extent of their contribution to land administration in Ekiti State remains understudied [31]. The research questions guiding this study include the following:

• What specific challenges hinder UAV adoption for cadastral mapping in Ekiti State, Nigeria?
• How does the accuracy of multi-resolution UAV imagery for cadastral boundary delineation vary across different settlement typologies?
• What are the practical implications and recommendations for leveraging UAVs to improve cadastral mapping and land administration in Ekiti State?

This study aims to investigate UAV adoption challenges; assess the use of multi-resolution UAV aerial imagery for land administration to identify their fitness for mapping cadastral boundaries in the various informal and formal contexts of rural, peri-urban, and urban areas; and propose strategies for leveraging UAVs. To address the gap between the potential of UAV technology and its practical implementation in cadastral mapping in the state and similar contexts, this study not only evaluates the technical accuracy of multi-resolution UAV imagery but also investigates the socio-economic factors influencing its adoption to provide a holistic approach for promoting the effective use of UAVs in land administration.

The next section presents an explanation of the data used and the rural, peri-urban, and urban areas mapped. It also contains the interviews conducted and the design of the survey carried out to identify factors leading to the limited adoption of UAVs for land administration and management in Nigeria. In Section 3, the results of the interviews, research survey, and aerial imagery survey are presented. Section 4 presents information on leveraging UAVs for cadastral mapping. It discusses the fitness of multi-resolution satellite imagery from perspectives of recognizability, settlement characteristics, and scalability. Finally, conclusions and recommendations about the adoption of UAVs and satellite imagery for cadastral mapping in rural, peri-urban, and urban contexts are made based on the findings obtained during this study (Section 5).

2. Materials and Methods

2.1. Study Area

Ekiti is one of the 36 states of Nigeria, aside from the Federal Capital Territory. Located within latitudes 7°16′N–8°7′N and longitudes 4°51′E–5°48′E, Ekiti lies at an average altitude of 528 m above the mean sea level. It is bordered by Kwara State in the north, Ondo State in the south and part of the southeast, Kogi State in the east, and Osun State in the west. With an approximate area of 5873 square kilometers [32], Ekiti accounts for only 0.57% of the total land mass of Nigeria, yet it is bigger in land mass when compared to Lagos, Anambra, and Abia States. Deriving its name from the Yoruba word for hills, i.e., Okiti, Ekiti is a significantly undulating terrain, having a height variation of about 474 m. Its lowest point is found toward the northeastern part of the state around Iye–Ekiti in the Ikole Local Government Area (LGA), with an elevation of 291 m of ellipsoidal height, and its highest point of 765 m is found around Ogotun–Ekiti in the Ekiti southwestern LGA. Ekiti is made up of several rural and urban communities, summing up to about 152 towns and villages. The 16 LGAs of the state are segmented into 3 senatorial districts, namely Ekiti North, Central, and South senatorial districts. The study area map of Nigeria, which shows the study sites of Ekiti and Ekiti States, is shown in Figure 1.

To understand how aerial imagery of various resolutions can support land administration across the various formal and informal contexts of rural, peri-urban, and urban landscapes, six settlement typologies were identified, which include the following: rural informal—Aaye-Oja; rural formal—Igedora; peri-urban informal—Aaye; peri-urban formal—Maryland Avenue; urban informal—Atinkankan; and urban formal—Okebola. The communities were selected based on their needs for UAV products and their suitability for the research. A seventh community—Egbewa Government Residential Area (GRA)—was mapped to highlight absolute positioning accuracy in the research. Aside from the above reasons, Atinkankan, Okebola, and Egbewa GRA were also mapped based on the need for orthophotos for planning purposes by the Ekiti State Geospatial Data Centre (ESGDC).

Table 1. Study area characteristics.

		Location	Type	Description
1	Rural	Part of Aaye-Oja Ekiti, Moba LGA	Informal	A mix of clustered rural settlements and large expanses of agricultural lands
2	Rural	Igedora Community, Irepodun/Ifelodun LGA	Formal	A low, densely populated community with some planning
3	Peri-urban	Aaye Community, Irepodun/Ifelodun LGA	Informal	Unplanned settlement with access to some basic amenities. Road, water, power
4	Peri-urban	Maryland Avenue, Ado-Ekiti, Ado LGA	Formal	Planned settlement with access to some basic amenities
5	Urban	Part of Atikankan, Ado-Ekiti, Ado LGA	Informal	A mix of urban slums with access to some basic facilities
6	Urban	Part of Okebola Street, Ado-Ekiti, Ado LGA	Formal	An urban area with some basic amenities
7	Urban	Part of Egbewa GRA	Formal	A Government Residential Area.

shows the study area characteristics.

Aaye-Oja Ekiti, situated in the Moba LGA of Ekiti State, is an ancient commercial community. Bordered by Otun and Igogo in the south, Ikosu in the east, Erinmope Ekiti in the north, and Osan in the west, Aaye-Oja is the nodal town of the Moba LGA of Ekiti State. Despite its nodal status and a comparable population among other towns in the LGA, Aaye-Oja lags in development compared to its counterparts of Otun and Erinmope-Ekiti. Its geometric properties are a mix of unplanned populated areas, scattered settlements, and expansive farmlands that characterize the landscape and make it a fitting representation of a rural area. This community typifies the rural informal settlements found in Ekiti State, highlighting a blend of traditional and underdeveloped features.

On the other hand, Igedora Community, a satellite town of Igede-Ekiti, the LGA headquarters of Irepodun/Ifelodun, stands as a rural settlement with well-planned layouts. Home to approximately 500 residents seeking the tranquility of rural living, Igedora combines the simplicity of rural life with organized architectural and planning designs. The community reflects a balance between rural serenity and planned infrastructure, highlighting an example of rural living with intentional community planning.

Aaye is a significantly homogeneous aboriginal settlement in the peri-urban expanse of Igede-Ekiti. It serves as an informal peri-urban settlement within the local government headquarters of the Irepodun/Ifelodun LGA. The community embodies the characteristics of an informal settlement, featuring a mix of traditional and irregular structures. Aaye community bridges the gap between rural simplicity and urban influence, presenting an environment that captures the essence of informal peri-urban space.

Maryland Avenue is a rapidly developing peri-urban area of Ado-Ekiti. It lies behind the Government Residential Area (GRA) 3rd extension. Maryland Avenue distinguishes itself with well-planned layouts as a formal, though peri-urban, settlement of Ado-Ekiti, the state capital of Ekiti State. Unlike typical peri-urban settings marked by informality, this community boasts organized structures and roads. The area benefits from its proximity to the GRA 3rd extension, highlighting the potential for planned expansion and urbanization.

Atinkankan is a commercial and residential hub in Ekiti State. It is a representation of an overpopulated urban area. Despite its centrality to Ado-Ekiti, the characteristics of Atinkankan align more with informal urban settlements. The area projects a slum-like settlement, reflecting the challenges associated with unplanned urban growth.

Okebola stands as a densely populated urban residential area in Ado-Ekiti, boasting essential amenities such as a government-owned school and a hospital. This densely populated urban setting represents a structured urban settlement with organized residential buildings. The area mapped in Okebola reflects the vibrancy and density associated with urban living, highlighting a blend of residential, institutional, and commercial establishments.

Egbewa GRA stands out as a planned settlement with relatively superior amenities compared to the other mapped settlements. We included it in the mapping specifically to facilitate a precise comparison between aerial imagery-derived data and GNSS data, ensuring absolute accuracy in our research findings.

2.2. Methods

This research employed a mixed-methods approach, combining interviews and surveys to assess socio-economic factors influencing UAV adoption with an evaluation of the precision of multi-resolution orthomosaics for boundary delineation. The survey data informed the selection of study sites and provided insights into the practical challenges faced by land professionals, while the accuracy assessment established the technical feasibility of UAV imagery for cadastral mapping.

To address Nigeria’s uniform landscape regarding UAV use and land administration laws, we designed and administered survey questions to uncover the factors influencing UAV adoption for land administration and management. We also conducted interviews with Nigerian professionals to gather valuable insights. The results were analyzed, and knowledge was drawn to identify pathways to leverage UAVs and aerial imagery for land administration and management in the country. Furthermore, through the lens of six distinct sites—namely parts of Aaye-Oja town, Igedora Community, Aaye Community, Maryland Avenue, Atinkankan, and Okebola—representing formal and informal rural, peri-urban, and urban landscapes in Ekiti State, Nigeria, potential cadastral boundaries were manually delineated through on-screen digitization using different resolution imagery of 0.05 m, 0.1 m, 0.5 m, and 1 m. Figure 2 illustrates the methodology flow. The inclusion of Egbewa GRA as the seventh settlement enabled a comparative analysis of the accuracy of 0.05 m and 0.1 m resolution imagery against GNSS-derived coordinates.

2.3. Factors Influencing Aerial Imagery Use

Employing a multifaceted approach, the study on UAV use incorporates perspectives through interviews and structured questionnaires to understand the use of aerial imagery for land administration in Nigeria, the factors contributing to the limited adoption of UAVs, and to identify potential solutions. Interviews with professionals in practice and government establishments were carried out to gain insight into Nigeria’s use of aerial imagery and UAVs for land administration and management-related projects. Three professionals actively involved in UAV mapping in Nigeria were interviewed. Their perspectives helped identify the current state of cadastral mapping with aerial imagery and UAVs in the country. This study utilized a quantitative research approach to investigate factors hindering UAV adoption for aerial imagery and devise effective adoption strategies for cadastral mapping. The questions targeted at surveyors, land administrators, and GIS analysts were developed and administered to participants using a random sampling approach. A total of 54 of the 60 respondents (90%) identified as surveyors, and 22 respondents (36.7%) identified as GIS specialists. Six respondents (10%) are land administrators, and 8.3% are spread across other professions. A link to the survey is provided in Appendix A. The survey was open for 30 days and was closed at the 60th unique response. Responses were categorized into thematic groups based on commonalities, and the frequency of each item was analyzed to determine recurring themes.

2.4. Multi-Resolution and Multi-Contextual Data Acquisition, Processing, and Analysis

2.4.1. Aerial Images Data Acquisition and Processing

UAV flights were conducted across the six sites ranging from rural to urban and informal to formal settlements using a DJI Mavic Pro Platinum UAV carrying a 10-megapixel resolution camera as part of a PhD research project. Three flights were conducted to cover part of Aaye-Oja. Single flights were conducted to cover parts of Igedora, Aaye community, Maryland avenue, Atinkankan and Okebola. Double flights were conducted to cover the area mapped in Egbewa GRA. All flights were conducted at 100 m flying height with forward and side overlaps of 80% and 72%, respectively. Agisoft Metashape Professional version 1.5.5 was used to process imagery of the six sites. Available 2.1 m panchromatic and 3.5 multispectral satellite imagery covering part of Aaye-Oja Ekiti, Ekiti State, was obtained from the National Space Research and Development Agency (NASRDA).

The orthomosaics produced from the processed UAV images were exported from Agisoft Metashape in the desired 0.05 m, 0.1 m, 0.5 m, and 1 m spatial resolutions. A 1-hectare rectangular shape defined on AutoCAD 2007 was selected across all six orthomosaics for equal comparison of study areas. Using the (GDAL) module, the Clip Raster by Extent tool of QGIS 2.14.17 was used to clip each image with the defined area. The 1-hectare area was selected to reflect the typical characteristics of the context for which the study area was chosen.

2.4.2. Reference Data for Accuracy Comparison

The absence of a reference cadastral data set necessitated the adoption of a relative accuracy comparison approach to maintain consistency across six-settlement typologies. Consequently, coordinates from the 0.05 m resolution image were designated as the reference, and coordinates from 0.1, 0.5, and 1 m imagery were compared to determine positional accuracy.

However, to show the benefits of reference data in the research, a seventh flight was conducted, dedicating additional time to acquiring GNSS data in Real-Time Kinematic (RTK) mode, which served as the reference. This enabled an absolute comparison with 0.05 and 0.1 orthomosaics, as the GNSS-coordinated marker points were not discernible in the lower-resolution 0.5 and 1 m imagery.

2.4.3. Multi-Resolution Aerial Imagery Analysis for Cadastral Mapping Fitness Determination

Four images from each site, representing different spatial resolutions of 0.05 m, 0.1 m, 0.5 m, and 1 m, were assessed for clarity and feature delineation at a 1:500 scale. Owing to the need to maintain consistency in the data sources and in the comparison within various settlement typologies, the 1-hectare area was clipped for visual assessment and coordinate comparison, coordinates of 10 points delineating parcels were identified from the digitized sets of coordinates, and 0.05 UAV imagery were adopted as reference. Horizontal Radial Root Mean Square Error (RMSEr) for the 0.1, 0.5, and 1 m resolution images were calculated. The area of a parcel was calculated from the delineated coordinates of each resolution of UAV orthomosaics (0.1, 0.5, and 1 m) and compared to the area of the parcel at 0.05 m resolution. The Normalized Parcel Area Error (NPAE), which expresses the relationship between the observed area of the parcel ($A_i$) at different image resolutions and the reference area of the parcel (${\widehat{A}}_i$) at the highest resolution of 0.05 m, was computed. This was computed by dividing the absolute difference between them by the reference area and multiplying the result by a normalizing factor (m), which was 1000 squared meters in this case. The equation allows us to compare the errors in parcel area estimation between the different image resolutions while accounting for the different scales and accuracies of the data.

The results of the RMSEr and NPAE across the various geographical contexts were analyzed and discussed to identify the fitness of the multi-resolution aerial imagery across multi-contextual geographical areas for cadastral mapping purposes.

RMSEr and NPAE were computed as follows:

$$\begin{gathered} RMSE= \sqrt{{\displaystyle \frac{1}{n}}\sum _{i=1}^n{((x_i-{\widehat{x}}_i)}^2+{((y_i-{\widehat{y}}_i)}^2)} \\ NPAE=\left({\displaystyle \frac{\left|A_i-{\widehat{A}}_i\right|}{{\widehat{A}}_i}}\right) \times m \end{gathered}$$

(1)

• $\mathrm{w}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{e} n \mathrm{i}\mathrm{s} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}.$
• $x_i \mathrm{a}\mathrm{n}\mathrm{d} y_{i }\mathrm{a}\mathrm{r}\mathrm{e} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d} x \mathrm{a}\mathrm{n}\mathrm{d} y \mathrm{c}\mathrm{o}\mathrm{o}\mathrm{r}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{s}, \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{l}\mathrm{y}.$
• ${\widehat{x}}_{i }\mathrm{a}\mathrm{n}\mathrm{d} {\widehat{y}}_{i }\mathrm{a}\mathrm{r}\mathrm{e} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{r}\mathrm{e}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e} x \mathrm{a}\mathrm{n}\mathrm{d} y \mathrm{c}\mathrm{o}\mathrm{o}\mathrm{r}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{s}, \mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{l}\mathrm{y}, \mathrm{f}\mathrm{o}\mathrm{r} 0.05 \mathrm{m} \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{h}\mathrm{i}\mathrm{s} \mathrm{c}\mathrm{a}\mathrm{s}\mathrm{e}$.
• $A_i \mathrm{a}\mathrm{n}\mathrm{d} {\widehat{A}}_{i } \mathrm{d}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{t}\mathrm{e} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a} \mathrm{s}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{r}\mathrm{e}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e} \mathrm{p}\mathrm{a}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{s}$.
• $m = Normalized Parcel Area Size \left(1000 {\mathrm{m}}^2\right) \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{h}\mathrm{i}\mathrm{s} \mathrm{c}\mathrm{a}\mathrm{s}\mathrm{e}$.

Also, the RMSEr approach adopted was repeated for the absolute comparison using GNSS data as reference, while the 0.05 m and 0.1 m imagery served as observed values.

3. Results

This section provides results of the interviews with professionals in practice and government in Nigeria; the survey was carried out with 60 respondents to obtain the assessment of multi-resolution orthomosaic imagery produced from the acquired UAV mapping process and the satellite imagery, the coordinates of the delineated points, and the areas of the delineated parcels.

3.1. The State of Land Administration with Aerial Imagery in Nigeria

From the interview conducted, it was realized that the use of aerial imagery and UAVs for land administration in Nigeria is a growing trend. However, challenges exist to the adoption and use of aerial imagery in the country. Edo State is a typical state that uses aerial imagery for land administration in the country. In the state, the government uses acquired aerial imagery as base maps to verify the consistency of existing survey plans produced by survey professionals before Certificates of Occupancies (C of O) are issued. When the need arises, some sites are visited and checked with GNSS receivers. As urbanization takes place, the need for re-surveying arises. The government, for land registration purposes, uses UAVs to update the base map and to prepare survey documents. While this approach is effective in most cases, it is laced with challenges when tree canopies cover demarcation points. When the fine details crucial for demarcation are not easily found on the image produced, traditional ground surveying methods are employed.

Other states, such as Lagos State, Sokoto, Gombe, Kaduna, Nassarawa, Benue, Plateau, the Federal Capital Territory, and Ekiti, are making efforts to incorporate aerial imagery into their cadastral mapping systems. Challenges referred to by the practitioners constraining the process in some of the contexts include the technical issues of non-overlapping images of contiguous areas and differences in datum parameters. Other challenges include vegetative cover limitation, difficulty in importing UAVs into Nigeria for professional surveying and mapping by non-governmental agencies and professionals, and the management of UAV imagery infrastructure. Both government and non-governmental bodies lamented the lack of a national regulatory framework for effective guidance of UAV surveys and the use of aerial imagery for cadastral mapping in Nigeria. The results of this interview helped to shape the design of the survey used to assess the factors affecting limited UAV use for land administration and management in Nigeria.

3.2. Factors Influencing UAV Use for Cadastral Mapping

Further insights into the factors influencing the use of UAVs for cadastral mapping in Nigeria and similar contexts were received in the survey conducted.

3.2.1. Population Characteristics

A total of 35% of respondents to the survey carried out to evaluate the “Factors Influencing Limited UAV Adoption for Land Administration and Management in Nigeria” are from Ekiti State, while the rest come from the other 35 states of Nigeria and the Federal Capital Territory (FCT).

Table 2. Survey population characteristics.

Variable	Options	Response
		Frequency (Multi-Choice)	Percentage
Occupation	Surveyor	54	90
	GIS Specialist	22	36.7
	Land Administrator	6	10
	Others	5	8.3
Practice Domain	Professional Practice	41	68.3
	Academic	15	25
	Government	9	15
	Others	4	6.7
Category	Surveying Student	3	5
	Young Surveyor	19	31.7
	Professional	40	66.7
	Others	4	6.7
Years of Experience	Less than 1 Year	2	3.3
	1–5 years	10	16.7
	6–10 years	23	38.3
	More than 10 Years	25	41.7

provides details of respondents to the survey. A total of 60 responses were received. From the survey result, 54.2% of respondents in Lagos State indicated the use of UAVs for land administration in Lagos State, while 47.7% acknowledged the same for the Federal Capital Territory. In both cases, results show that the implementations face some challenges.

3.2.2. Awareness and Education

A total of 53.3% (32) of respondents to the survey agreed to have used aerial imagery for cadastral mapping in Nigeria, 55% (33) of the respondents claimed to be very aware of the use of UAVs and their applications in land administration, 38.3% (23) claim to be moderately aware, and 6.7% (4) claim to be unaware. While accessing the number of respondents that have received formal education or training, only 56.7% (34) of respondents indicated to have received education from secondary or tertiary institutions, mandatory continuous professional development programs, or other UAV certification training. This underscores the importance of increasing accessibility to UAV training among current and prospective land professionals in the country.

3.2.3. Perceived Benefits and Concerns of Using UAVs for Land Administration

In response to the question that asked of the potential benefits associated with the use of UAVs in land administration, time efficiency ranks the highest, followed by enhanced data resolution. Improved accuracy ranks the lowest, with 53.3% acknowledging it. This implies that the population sampled generally accepts UAVs as beneficial to land administration. Respondents suggested other benefits of UAVs, such as confidence and reliability, access to inaccessible locations, data storage and retrieval, and security and safety in volatile areas. Figure 3 illustrates the quantitative result.

3.2.4. Concerns and Challenges Perceived in Adopting UAVs for Land Administration and Management in Nigeria

Among the challenges perceived in adopting UAVs for land administration and management in Nigeria, the cost of UAV technology ranks the highest at 68.3% (41), and data security concerns rank the lowest at 33.3% (20). Respondents also opined that a lack of regulatory framework and limited expertise are major challenges in adopting UAVs for land administration. This result is displayed in a horizontal bar graph in Figure 4. Other challenges outlined by respondents include bottlenecks in acquiring the necessary certification for UAV operation from regulatory agencies, acceptability issues, and UAV pilot licensing as the factors militating against the adoption of UAVs for land administration and its wider land management functions in the country.

3.2.5. Regulatory and Policy Environment

A total of 58.3% (35) of the respondents indicated not being aware of specific regulations guiding the responsible use of UAVs in land administration in their state. This corroborates the lack of regulatory framework mentioned by professionals in the interviews conducted.

A total of 6.7% (4) of the respondents rated the regulatory framework as excellent, and 20% (12) of the respondents considered the regulations good, acknowledging their adequacy but suggesting room for minor improvements; 25% (15) of respondents expressed that the regulatory framework had some gaps or inconsistencies, indicating a need for improvements to enhance its effectiveness; 16.7% (10) of respondents indicated substantial deficiencies in the existing regulations, suggesting that significant reforms are required for effectiveness; while 31.7% (19) of respondents abstained from expressing an opinion, citing insufficient knowledge or experience to make a judgment. This further corroborates the need for improved regulations and documents to guide the use of UAVs across the countries.

In identifying the specific regulatory challenges hindering the widespread adoption of UAVs for land administration and management in Nigeria, respondents identified several challenges having the below listed as most re-occurring:

• Licensing and certification: Participants highlighted challenges related to obtaining licenses and certifications for UAV operation. Issues such as inconsistent licensing procedures and the lengthy process of obtaining the End User Certificate (EUC) were prominent concerns;
• Security and regulatory restrictions: Respondents expressed concerns about security issues, particularly the identification of no-fly zones, restrictions imposed by military personnel (e.g., the Nigerian Army), and limitations due to airport or air force authority regulations;
• Government policies and awareness: Participants cited challenges related to government policies, funds, and the absence of enabling laws for UAV use. Additionally, the lack of awareness among stakeholders was identified as a barrier to widespread adoption.

3.2.6. Cost and Accessibility

Cost-related challenges were a recurring theme, including the cost of UAV technology, the expenses associated with obtaining necessary licenses, and the financial requirements for satisfying regulatory procedures. Results indicate that the cost associated with acquiring and maintaining UAV technology is a barrier to the use of the technology in the Nigerian context. In total, 16 (26.7%) and 27 (45%) of the 60 respondents strongly agree and agree, respectively, that the cost of acquiring and maintaining UAV technology is a barrier to leveraging the technology in the Nigerian Context. Furthermore, 28 of the 60 respondents provided additional information on the cost and accessibility of UAV technology in Nigeria. The most frequently cited challenge was the high cost of acquiring UAVs and the associated software for image processing. Free and open-source software, such as OpenDroneMap, WebODM, etc., were identified as alternatives for UAV data processing. Additional concerns included the high cost of licenses and certificates for UAV pilots, as well as the financial implications of data management. Participants emphasized the overall high cost attributed to the importation of the equipment.

In addition, a larger percentage of participants opined on the non-availability of technical infrastructure to drive UAV technology in the country, as 6.7% (4) and 58.3% (35) responded that it is insufficient and sufficient, respectively. Participants underscored the need for investments to enhance the technical infrastructure for UAV adoption.

3.2.7. Leveraging UAV Technology for Aerial Imagery Provision in Nigeria

In leveraging UAV technology for land administration, participants stressed the need for government and private support for training institutions, integration of UAVs into professional bodies’ continuing education programs and seminars, and the establishment of a regulatory framework. Financial support and government policies were highlighted as crucial factors to promote the acceptance and affordability of UAVs for mapping purposes. Recommendations extended to creating a UAV board for pilot certification, increasing awareness through campaigns, and integrating UAV technology into higher education curricula.

Responses from participants underscore significant gaps in UAV training, particularly the absence of a formal curriculum devoted to UAV training in Nigerian higher learning institutions. The suggestions focused on the imperative need for comprehensive UAV training programs in polytechnics, universities, and survey-related institutions. Despite the current limitations, participants expressed optimism about the evolving survey methods, transitioning from analog to digital and now incorporating visual technology. It is believed that widespread UAV adoption is imminent. Participants stressed the importance of disseminating research results to generate interest. Additional comments highlight the critical role of UAVs in addressing security challenges in Nigeria and emphasize the need for enhanced awareness regarding their ethical use.

3.3. Analysis of UAV and Satellite Multi-Resolution Imagery in Varied Geographical Contexts for Cadastral Mapping

3.3.1. Differences in Mapping Rural, Peri-Urban, and Urban Areas

From the UAV flights, imagery of 2.16, 3.39, 3.02, 2.88, 3.19, and 3.20 cm resolution covering Aaye-Oja, Igedora, Aaye Community, Maryland Avenue, Atinkankan, and Okebola, respectively, was produced. The 1-hectare areas clipped out of each image were exported at 0.05, 0.1, 0.5, and 1 m resolution. Figure 5 shows the orthomosaic map of part of Aaye-Oja Ekiti.

The orthomosaic imagery of part of Aaye-Oja Ekiti exhibits characteristics typical of rural and informal settlements with linear or clustered settlement patterns and large expanses of farmlands. This particular image portrays a mix of unplanned and scattered structures interspersed with farmland. Buildings are irregularly placed, and property boundaries are undefined. The lack of formal planning is evident in the absence of structured layouts and organized demarcations. Being a rural area, most properties do not have physical demarcations.

The imagery of Igedora (Figure 6) presents the geometric characteristics of a rural and formal settlement. The landscape shows well-planned layouts with organized, distantly placed building structures. The geometric features include clearly defined property boundaries, structured road networks, and a planned arrangement of buildings. The UAV aerial imagery reflects a form of formalized rural settlement.

Aaye community’s UAV aerial imagery depicts an informal peri-urban settlement (Figure 7). The geometric characteristics reveal a mix of unplanned structures, scattered settlements, and an absence of formal organization. The landscape lacks the well-defined boundaries and planned layouts seen in formal settlements, displaying the informality typical of peri-urban areas.

In Maryland Avenue, the aerial imagery captures the well-laid-out buildings with structured road networks. Property boundaries are clearly defined with fences, except for the open spaces that are prevalent owing to the gradual transition from rural to urban areas (Figure 8).

Imagery from Atinkankan reveals the geometric characteristics of an urban and informal settlement (Figure 9). The northwestern landscape shows densely packed, irregularly shaped slum-like settlements with limited infrastructure. It presents a lack of organized layouts and undefined property boundaries that typify informal urban development.

On the other hand, the imagery of the area mapped in Okebola exhibits the characteristics of an urbanized area (Figure 10). The landscape features well-laid-out buildings, structured road networks, and clearly defined property boundaries. The area typifies a formalized and regulated urban development with organized layouts.

The orthomosaic of Egbewa GRA reveals an urban landscape characterized by neatly arranged buildings, systematic road networks, and distinct property boundaries. This area embodies a formally planned and regulated urban environment, marked by orderly layouts and well-defined spatial organization, as shown in Figure 11.

3.3.2. Interpretability of Cadastral Boundaries from Multi-Resolution Imagery

The interpretability of cadastral boundaries from the 0.05, 0.1, 0.5, and 1 m resolution aerial imagery of the same site was examined by visually comparing recognizable features across the six study sites at a 1:500 visualization scale on QGIS 2.14.17, as shown in Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17. In all cases, the problem of tree canopies obstructing the delineation process from the computer with aerial imagery persists. Figure 12 represents the imagery of a typical rural agricultural area, showing small farms extracted from the rural informal settlement of Aaye-Oja. In the 1 m resolution image of the area Figure 12a, individual farms can only be identified. The image lacks appropriate fineness to aid the recognition of boundary demarcation points. However, at the higher resolutions of 0.5, 0.1, and 0.05, boundary points can be identified and delineated Appendix B.

Typical of formal settlements, the rural area of Igedora is significantly laced with fences as physical boundaries, thereby making property demarcation easier not only for the people having rights to the property but also for third persons. From Figure 13a with 1 m resolution, although features like buildings and farmlands are recognizable, the image is not refined enough for clear demarcation. On the other hand, property boundaries are easily identified in Figure 13b–d, which corresponds to 0.5, 0.1, and 0.05 m resolutions, respectively. At the 1:500 zoom level adopted, the difference between Figure 13c,d cannot be appreciated. The difference in 0.1 and 0.05 m resolution imagery only became visually significant at a scale of 1:250 on the QGIS 2.14.17 used.

Figure 14 represents a clip from part of the peri-urban informal settlement of the Aaye community. Features like buildings, roads, and vegetation can be recognized in the 1 m resolution image in Figure 14a. However, the image lacks adequate fineness to aid proper delineation or identification of smaller features like wells, boreholes, etc. On the other hand, Figure 14b gives better clarity to features. Tree types are easily identified with higher-resolution imagery. The difference between Figure 14c and d remains unclear when observed visually at a scale lower than 1:250 on the QGIS 2.14.17 map canvas.

Figure 15 is a clip from the typical peri-urban formal settlement. Because of the prevalence of fences as the physical demarcation between parcels, boundaries can be perceived in the 1 m resolution image, as shown in Figure 13a. However, recognizing demarcation lines in vegetated areas is difficult. In Figure 13b with 1 m resolution, some physical demarcation lines were observed in the vegetated areas, but this is insufficient in demarcating peri-urban areas with the level of accuracy required. Identifying and demarcating boundaries, especially fences, come with ease in Figure 15c. Non-linear features such as water tanks and wells that often lie close to boundaries were also easily identified and can aid in the demarcation process of properties in this context. The difference between Figure 15c at 0.1 m resolution and Figure 15d at 0.05 m resolution remains unperceivable with visual examination at this scale.

Figure 16 is a clip from the peri-urban informal settlement of Atinkankan. The clip presents a typical slum within the urban context containing small and closely packed buildings. The Figure 16a 1 m resolution image is blurry, and the non-presence of physical demarcation that characterizes the area necessitates alternative approaches from using the 1 m resolution image difficult for identifying property boundaries. In addition, fences and other boundary demarcation objects remain uneasily recognizable at the 0.5 m resolution image of Figure 16b. Footpaths, building edges, and other delineating features are easily perceived in Figure 16c at 0.1 m resolution. This highlights the need for at least a 0.1 m resolution image to aid aerial imagery delineation in slum-like urban centers. At the scale of observation, no clear difference was observed between resolutions of 0.1 and 0.05 m.

Figure 17 is a clip from Okebola, representing formal urban settlement. The patterned arrangement makes boundary lines easily discernible. Although images of 1 m and 0.5 m resolution are blurry, the regular pattern makes recognizing linear boundary patterns possible. In Figure 17a, boundaries appear blurry but recognizable. However, the required level of accuracy for demarcating properties in an urban context and the potential risk of inaccurate information, owing to the value and quest for land in the area, makes it difficult to use a 1 m resolution image for boundary delineation. Recognizing features became better with the 0.5 m resolution image in Figure 17b. However, smaller point features like wells, electric poles, etc., are still not quite visible. In Figure 17c, recognizing boundary demarcation features came with ease. Differences were not observed between 0.1 m and 0.05 m resolution imagery at the scale adopted for visual perception.

In orthomosaics, it is easier to identify features such as holes between fences and pillars at higher resolutions of 0.1 and 0.05 m. In 0.5 m imagery, it becomes difficult to distinguish features like beacons or holes in fences, but patterns and intersections of linear features, especially fences, help to identify demarcation points. For 1 m imagery, most demarcation features are not easily identified, thereby hampering boundary delineation processes. At this resolution, delineation is possible only while working at a much smaller scale, which often makes the digitization less accurate. Ultimately, the recognizability of physical boundaries in aerial imagery aids the correct delineation of features for cadastral mapping purposes.

3.3.3. Differences in Delineating Cadastral Boundaries at Various Resolutions

RMSEr is the error or discrepancy between observed values and reference values. The RMSEr gives a sense of overall error magnitude, with larger values indicating potentially poorer accuracy. Using the coordinates extracted from the 0.05 m resolution imagery as a reference, the RMSEr of points delineated from 0.1, 0.5, and 1 m resolution imagery was calculated for Aaye-Oja, Igedora, Aaye, Maryland, Atinkankan, and Okebola.

Table 3. RMSEr of points delineated from multi-resolution UAV imagery.

Site	Description	Root Mean Square Error (RMSEr)
		0.1 m	0.5 m	1 m
Aaye-Oja	General boundaries from informal settlements comprising road intersections and farm edges	0.190	0.613	2.572
Igedora	Mix of physical and general boundaries determined from fences and farm edges	0.142	0.740	2.501
Aaye	Mix of physical and general boundaries determined from fences, road, and farm edges	0.075	0.680	1.543
Maryland	Physical boundaries from fences	0.053	1.332	1.417
Atinkankan	Physical and general boundaries extracted from fences and culvert edges	0.157	0.570	1.491
Okebola	Physical boundaries from fences	0.106	0.535	1.121
	Average RMSEr values	0.117	0.737	1.747

shows the characteristics of boundaries observed and the RMSEr obtained from the 10 points delineated within the 1-hectare area clipped from the six contexts.

The tables showing the RMSEr computed from each observation before it was aggregated in Table 3 are shown in the Appendix C section.

The absolute comparison conducted at Egbewa GRA using RTK GNSS-derived coordinates as reference yielded RMSEr values of 0.194 m and 0.206 m for 0.05 m and 0.1 m aerial imagery. The GNSS observed coordinates and coordinates from multi-resolution orthomosaics are contained in

Table 4. RMSEr between GNSS-derived coordinates and 0.05 m resolution orthomosaic.

	GNSS Survey		0.05 Orthomosaic
Point ID	Northings (m)	Eastings (m)	Northings (m)	Eastings (m)	N0.05–NGNSS (m)	E0.05–EGNSS (m)	(N0.05–NGNSS)2 (m2)	(E0.05–EGNSS)2 (m2)
PT1	843,316.647	743,574.247	843,316.685	743,574.355	0.038	0.108	0.001	0.012
PT1A	843,294.853	743,577.607	843,294.833	743,577.497	−0.020	−0.110	0.000	0.012
PT1B	843,339.240	743,572.166	843,339.191	743,572.457	−0.049	0.291	0.002	0.085
PT2	843,294.353	743,854.335	843,294.354	743,854.243	0.001	−0.092	0.000	0.009
PT2B	843,280.579	743,881.948	843,280.744	743,881.692	0.165	−0.256	0.027	0.066
PT2C	843,301.188	743,828.457	843,301.198	743,828.436	0.010	−0.021	0.000	0.000
PT5A	843,408.008	743,795.789	843,407.863	743,795.802	−0.145	0.013	0.021	0.000
PT5B	843,427.010	743,674.443	843,427.006	743,674.616	−0.004	0.173	0.000	0.030
PT6	843,873.704	743,576.325	843,873.817	743,576.238	0.113	−0.087	0.013	0.008
PT6A	843,881.845	743,555.280	843,881.930	743,555.405	0.085	0.125	0.007	0.016
PT6B	843,864.870	743,599.227	843,865.023	743,599.145	0.153	−0.082	0.023	0.007
PT8	843,733.918	743,798.118	843,733.633	743,797.940	−0.285	−0.178	0.081	0.032
							0.122	0.151
					0.05 RMSEr:		0.194

and

Table 5. RMSEr between GNSS-derived coordinates and 0.1 m resolution orthomosaic.

	GNSS Survey		0.1 Orthomosaic
Point ID	Northings (m)	Eastings (m)	Northings (m)	Eastings (m)	N0.1–NGNSS (m)	E0.1–EGNSS (m)	(N0.1–NGNSS)2 (m2)	(E0.1–EGNSS)2 (m2)
PT1	843,316.647	743,574.247	843,316.6722	743,574.3841	0.025	0.137	0.001	0.019
PT1A	843,294.853	743,577.607	843,294.8272	743,577.4969	−0.026	−0.110	0.001	0.012
PT1B	843,339.240	743,572.166	843,339.1871	743,572.4427	−0.053	0.277	0.003	0.077
PT2	843,294.353	743,854.335	843,294.3714	743,854.2452	0.018	−0.090	0.000	0.008
PT2B	843,280.579	743,881.948	843,280.7391	743,881.697	0.160	−0.251	0.026	0.063
PT2C	843,301.188	743,828.457	843,301.1808	743,828.3991	−0.007	−0.058	0.000	0.003
PT5A	843,408.008	743,795.789	843,407.8339	743,795.8027	−0.174	0.014	0.030	0.000
PT5B	843,427.010	743,674.443	843,426.9763	743,674.6424	−0.034	0.199	0.001	0.040
PT6	843,873.704	743,576.325	843,873.8234	743,576.1993	0.119	−0.126	0.014	0.016
PT6A	843,881.845	743,555.280	843,881.9251	743,555.3949	0.080	0.115	0.006	0.013
PT6B	843,864.870	743,599.227	843,865.026	743,599.0956	0.156	−0.131	0.024	0.017
PT8	843,733.918	743,798.118	843,733.6282	743,797.8981	−0.290	−0.220	0.084	0.048
							0.126	0.162
					0.1 RMSEr:		0.206

. Notably, these values exhibit a significant increase in RMSEr compared to the relative comparison values, highlighting the inaccuracies introduced by comparing data from two distinct observation methods (GNSS and aerial imagery). In addition, the impact of PT1B, PT2B, and PT8, which could be regarded as outliers, could have contributed to the significant discrepancy. However, when the three data were excluded from the RMSEr calculations, 0.133 m and 0.152 m RMSEr values were obtained, which makes the values closer to the relative comparison conducted. This discrepancy emphasizes the importance of considering the differences in measurement techniques when evaluating positional accuracy.

The horizontal accuracy of the orthomosaics meets the American Society for Photogrammetry and Remote Sensing ASPRS (2014) Positional Accuracy Standards for Digital Geospatial Data, with the 0.05 m GSD achieving 0.194 m RMSEr (Class III), suitable for general mapping and visualization with moderate to low accuracy. In contrast, the 0.1 m GSD attains 0.206 m RMSEr (Class II), indicating standard and high-accuracy mapping-grade geospatial data. Notably, the average RMSEr value of 0.133 m for seven checkpoints on the 0.05 m orthomosaic conforms to Class II, while the 0.1 m GSD is marginally below Class I. Both Class I and Class II accuracy levels are suitable for cadastral mapping, depending on settlement typology and local requirements. Table 4 and Table 5 present the comparison results between GNSS-derived coordinates and orthomosaics with resolutions of 0.05 m and 0.1 m, respectively.

In addition, the result of the NPAE computed to show the effects of the coordinates delineated from different resolutions of imagery on the sizes of the parcels is shown in

Table 6. NPAE from multi-resolution imagery in various geographical contexts.

Site	Area in Squared Meters (m2)				Absolute Error in Squared Meters (m2)			Absolute Error Normalized to 1000 Square Meters (m2)
	0.05 m	0.1 m	0.5 m	1 m	0.1	0.5	1	0.1	0.5	1
Aaye-Oja	994.028	984.066	986.359	1024.062	9.962	7.670	30.034	10.022	7.716	30.214
Igedora	726.928	728.452	747.314	796.060	1.524	20.386	69.133	2.097	28.044	95.102
Aaye	888.676	890.272	908.521	869.699	1.596	19.845	18.978	1.796	22.331	21.355
Maryland	1056.242	1054.463	1064.629	1072.923	1.779	8.387	16.681	1.684	7.940	15.793
Atinkankan	111.578	109.441	100.784	110.586	2.136	10.794	0.991	19.148	96.737	8.884
Okebola	1026.079	1027.764	1037.476	1079.781	1.685	11.396	53.702	1.642	11.107	52.337

. The cross-coordinate procedure adopted for parcel area calculation is also shown in the Appendix C section of the report.

Lower values of RMSEr indicate better accuracy and precision. Generally, as image resolution decreases from 0.1 m to 1 m, the RMSEr values tend to increase. This is because lower resolutions result in less detailed images, leading to less accurate measurements. However, some sites exhibit more sensitivity to resolution changes than others do, depending on the complexity of the features present within them. Sites with more complex features or boundaries, such as Aaye-Oja and Igedora, tend to have higher RMSEr values compared to sites with simpler features, such as Okebola.

Informal settlements often lack formal planning, leading to irregular and complex boundaries that are non-easily discernible. The higher RMSEr values for Aaye-Oja and Atinkankan could be attributed to the complexity of features such as road intersections, farm edges, and culvert edges, which are less discernible when compared to more organized and well-defined physical boundaries such as fences, which results in lower RMSEr typical of Okebola and Maryland.

The absolute errors represent the discrepancies between the observed parcel sizes and the reference sizes of 0.05 m², with lower values indicating greater positional accuracy. The absolute errors of different sizes observed across 0.1, 0.5, and 1 m resolution imagery were normalized to 1000 m² to allow for a fair comparison. The trend observed across formal geographical contexts, where fences served as physical demarcations aiding easy delineation, was more consistent than that observed across informal contexts. The 0.1 m resolution had a mean normalized absolute error of 1.807 m² in the formal context, while the 1 m resolution had the highest inaccuracy.

3.3.4. Differences in UAV Orthomosaic and Satellite Imagery for Cadastral Mapping

Freely available 2.1 and 3.5 m resolution satellite imagery acquired from NASRDA were compared with the high-resolution orthomosaic aerial imagery produced from the UAV flights for cadastral mapping purposes. The results agree with existing research that UAV orthomosaics typically have a much higher spatial resolution than satellite imagery, allowing for more detailed and accurate mapping [19]. UAV orthomosaics generally provide spatially explicit detail compared to satellite imagery. UAV imagery is subjected to lesser levels of atmospheric distortions than satellite imagery. UAV imagery also allows for the high temporal resolution required for cadastral map updating [22,24,30]. Maps made from the 2.1 and 3.5 m resolution satellite imagery clipped at 1:20,000 visualization scale of QGIS 2.14.17 for Aaye-Oja Ekiti are shown in Figure 18. Large farms are visible on both imagery but more visible on the 2.1 m panchromatic imagery than on the 3.6 multispectral image. However, the boundaries of the farms remain unclear because of the low spatial resolution of the imagery. In addition, small farms, such as those visible in Figure 12b–d, cannot be seen in the satellite imagery.

As found in the existing body of knowledge, UAV orthomosaics can be captured more frequently than satellite imagery, providing up-to-date information for cadastral mapping. UAVs are cost-effective for mapping small areas, while satellite imagery is more cost-effective for mapping large areas. However, with the development of fixed-wing UAVs with the ability to take longer flights, UAVs are becoming increasingly cost-effective for mapping large areas compared to buying satellite imagery data acquisition. UAV orthomosaics can be produced for specific areas of interest, while satellite imagery typically covers larger areas. UAV orthomosaics can be captured and processed more quickly than satellite imagery, providing faster access to information for cadastral mapping [19,24]. UAV flights are restricted in certain places, while satellite imagery is easily utilized without restrictions in various applications. The results are tabulated for clarity in

Table 7. Comparison between UAV orthomosaics and satellite imagery for cadastral mapping.

Factors	UAV	Satellite Imagery
Resolution	Higher resolutions (0.01 m) can be achieved	Usually lower resolutions
Area	Usually for small areas	Wide coverage aids synoptic view
Cost	Cost-effective for small areas	Cost-effective for large areas
Regulations	Restrictions, especially in urban areas	No restrictions
Flexibility	Ease of data collection aids high-temporal resolution for real-time data collection	Low-temporal resolution

.

3.3.5. Cadastral Mapping Resolution Guide

Results from this study inform the development of a guideline for selecting optimal imagery resolution for cadastral mapping

Table 8. Cadastral mapping resolution guide.

Context	Settlement Example	Resolution
Rural Informal	Aaye-Oja	0.5 m (0.1 m in dense areas)
Rural Formal	Igedora	0.5 m (0.1 m in dense areas)
Peri-Urban Informal	Aaye	0.1 m
Peri-Urban Formal	Maryland Avenue	0.1 m
Urban Informal	Atinkankan	0.1 m–0.05 m
Urban Formal	Okebola	0.1 m–0.05 m

. The minimum resolution requirements will vary based on local conditions, land use patterns, and desired map detail.

The use of point or line features for boundary delineation significantly influences clarity. Orthomosaics with prominent line features, such as fences and roads, facilitate easier delineation compared to those relying on point features like pillars. Line features enhance the visibility of demarcation points.

In FFPLA contexts, where coverage prioritizes conformance over technical standards, affordability should guide resolution selection. Higher resolutions increase data acquisition and processing costs. In certain cases, combining multiple resolutions may be necessary to capture both large-scale context and fine-scale details within settlements, particularly in rural areas with densely clustered buildings at the core.

4. Discussions

The findings of the socio-economic survey revealed that cost, regulatory barriers, and lack of awareness are major obstacles to UAV adoption in Nigeria. These factors highlight the importance of not only demonstrating the technical accuracy of UAV imagery, as evidenced by the RMSEr and NPAE results but also addressing the broader socio-economic context to ensure its successful integration into land administration practices.

4.1. Leveraging UAVs for Cadastral Mapping

The results of the interview and survey provided insights into the factors influencing the adoption of UAVs for cadastral mapping in Nigeria. Key findings from the result include the following:

• Awareness and education: The high level of awareness among respondents (55% very aware, 38.3% moderately aware) suggests that there is a growing recognition of the potential of UAVs in land administration. However, the relatively lower percentage of respondents who received formal education or training in UAVs (56.7%) indicates a need for increased accessibility to formal training programs;
• Perceived benefits: The perceived benefits of UAVs, particularly in terms of time efficiency and enhanced data resolution, are consistent with previous studies [24]. These findings reinforce the notion that UAVs offer significant advantages over traditional methods of cadastral mapping;
• Concerns and challenges: The concerns identified by respondents include the cost of UAV technology and regulatory challenges. These findings highlight the need for policymakers and industry stakeholders to address these concerns in order to facilitate wider adoption of UAVs. The recurring challenges identified by respondents, such as licensing and certification, security and regulatory restrictions, cost factors, knowledge and expertise, and government policies, are consistent with previous studies [21,24]. These challenges pose significant barriers to the adoption of UAVs and need to be addressed in order to unlock the full potential of this technology;
• Advocacy: Stakeholders need to advocate for UAV regulatory revisions and the adoption of UAVs as a source of remotely sensed data to facilitate large-scale cadastral mapping.

4.2. Fitness of Multi-Resolution Aerial Imagery for Cadastral Mapping

The delineation of features significantly depends on the availability and recognizability of physical objects that delineate the parcels based on the imagery produced. This study considered varying resolutions of aerial imagery and compared the outcomes across rural, peri-urban, and urban areas, shedding light on the suitability of aerial imagery for cadastral mapping. Considering the settlement characteristics, resolution impact on mapping, image size, and computational implications, and the differences between UAV orthomosaics and satellite imagery, this study drew conclusions on the fitness of UAVs and Aerial Imagery for cadastral mapping in Nigeria and similar developing countries. The visual representations in Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10 and Figure 11 highlight the distinct characteristics of the settlements mapped.

The analysis of aerial imagery at resolutions of 0.05 m, 0.1 m, 0.5 m, and 1 m provides crucial insights into the fitness of different resolutions for mapping purposes. The comparisons demonstrated in Figure 12, Figure 13, Figure 14, Figure 15, Figure 16, Figure 17 and Figure 18 show that as resolution increases, the clarity of features improves, making finer details and boundary demarcations more discernible. This is particularly important for accurately delineating property boundaries, identifying features like wells and electric poles, and ensuring precise mapping in various settlement types. The analysis of RMSEr and NPAE also provides important insights into the suitability of the multi-resolution aerial imagery for cadastral mapping in multi-contextual geographical scenarios.

The comparison between UAV orthomosaics and satellite imagery revealed the advantages of UAV technology for cadastral mapping. UAV orthomosaics exhibited higher resolution, allowing for more detailed and accurate mapping. The UAV data also offered more frequent updates, cost-effectiveness for smaller areas, and faster data collection and processing. However, satellite imagery of between 0.5 and 0.1 m was deemed more cost-effective for large-area mapping.

4.3. Practical Implications and Recommendations

• Physical demarcations are fundamental for cadastral boundary delineation: In the context of general boundaries and physical boundaries for and beyond FFPLA cadastral mapping, discernibility of physical demarcations is vital in achieving accurate and precise coordinates required for avoiding conflicts;
• Settlement typology awareness: This study emphasizes the importance of considering settlement characteristics when choosing the appropriate resolution for cadastral mapping. Different resolutions are more suitable for different settlement types. Rural areas are dominated by farmlands. Hence, the use of lower-resolution imagery, like the 0.5 m resolution imagery, can suffice. However, higher-resolution imagery is required for mapping urban areas because of the dense nature of properties and the need for more accurate spatial data;
• UAV advantages for cadastral mapping: This study underscores the advantages of UAV orthomosaics over satellite imagery in terms of resolution, accuracy, and flexibility. This necessitates the creation of pathways for the sustainable use of UAVs for professional purposes;
• Regulatory considerations: This study hints at restrictions on UAV flights for professionals, emphasizing the need for policymakers to address regulatory challenges hindering UAV technology’s widespread use by professionals. Professionals should be licensed and allowed to import and use UAVs responsibly when needed.

For land administration purposes, where the need to fast-track land registration has become urgent, leveraging 0.5 m image resolution to minimize cost and maximize efficiency becomes very vital.

4.4. Limitations and Recommendations for Further Studies

This study aimed to assess the fitness of multi-resolution remotely sensed data for cadastral mapping in both informal and formal contexts within rural, peri-urban, and urban areas. This work focuses on human delineation as opposed to automatic feature extraction procedures. Specifically, the investigation focused on UAV imagery at resolutions of 0.05 m, 0.1 m, 0.5 m, and 1 m, along with satellite imagery at 2.1 m and 3.5 m. Emphasis was placed on the identification of fences and other linear features compared to point features. The identification of beacons was not the primary focus, even though 40 × 40 cm second-order control beacons were identified on the 0.1 m and higher-resolution images. In addition, the presence of trees impedes easy boundary demarcation with aerial imagery, necessitating field visits in such circumstances. This study found no significant differences in human identification of boundaries between 0.1 m and 0.05 m aerial imagery resolutions at the 1:500 visualization scale. However, when observed at a larger scale of 1:125 during delineation, some distinctions became apparent between imagery with 0.1 m and 0.05 m resolutions.

While this study investigates UAV adoption challenges in Nigeria, based on the context of the research, it is recommended that future research should investigate stakeholders’ willingness to adopt UAV technology and their perceived barriers to change. Additionally, a comprehensive evaluation of the regulatory framework governing UAV use is needed to identify necessary revisions that can facilitate the integration of UAVs for cadastral mapping. Addressing these knowledge gaps will help to promote its effective use. Some other suggestions for further studies are advanced:

• Automatic feature extraction: Explore the potential of using machine learning and computer vision algorithms for the automatic extraction of cadastral boundaries and other relevant features from UAV imagery in diverse contextual scenarios. This could significantly improve the efficiency and accuracy of cadastral mapping, especially in large and complex areas;
• Integration of UAVs and GNSS: Investigate the optimal integration of UAVs and GNSS technologies for cadastral mapping, particularly in terms of data fusion, accuracy assessment, and workflow optimization. This could lead to the development of hybrid mapping approaches that leverage the strengths of both technologies;
• Impact on land governance: Evaluate the long-term impact of UAV-based cadastral mapping on land governance and socio-economic development, particularly in terms of land tenure security, dispute resolution, and access to land-related services. This would help to understand the broader benefits of the technology beyond mapping accuracy;
• Comparison with other remote sensing technologies: Compare the performance of aerial imagery with Light Detection and Ranging (LiDAR) methods for cadastral mapping in different contexts. This would provide a comprehensive understanding of the strengths and limitations of each technology and guide their selection for specific applications;
• Capacity building and training: Evaluate how the implementation of formal capacity building and training programs for land professionals and other stakeholders on the use of UAVs for cadastral mapping would impact the acceptability or otherwise of UAVs and other remotely sensed data for cadastral mapping.

5. Conclusions

This paper aimed to identify some factors affecting the adoption of UAVs for cadastral mapping in land administration. It also provides an analysis of the use of aerial imagery with multi-resolutions ranging from 0.05 m to 1 m for boundary delineation in diverse rural, peri-urban, and urban multi-contextual geographical areas using RMSEr and NPAE. By integrating insights from both the socio-economic survey and the accuracy assessment, this study provides insights into the fitness of remotely sensed data for cadastral mapping in multi-contextual scenarios. This study contributes to UAV applications in cadastral mapping and aerial imagery use for cadastral mapping, emphasizing resolution considerations, settlement characteristics, and the practical adoption of UAVs versus satellite imagery. The findings highlight the importance of selecting appropriate imagery resolutions based on settlement typologies, recognizing that different resolutions suit various settlement types.

This study emphasizes the advantages of UAV orthomosaics over satellite imagery in terms of resolution, accuracy, and flexibility. This underlines the importance of fostering pathways for sustainable UAV use in professional land administration practices; by addressing these aspects, this paper contributes to improved efficiency in spatial data gathering and enhanced land governance practices using multi-resolution UAVs and satellite aerial imagery.