**1. Introduction**

An unmanned aerial vehicle (UAV) is an unmanned aircraft operated by radio remote control equipment and self-provided program control device [1]. The combination of a UAV and a remote sensing sensor can constitute an ultralow-altitude remote sensing monitoring system. UAV remote sensing has many advantages, such as high image spatial resolution, instant information acquisition, convenient operation, high maneuverability, freedom from cloud interference, and low cost [2–4]. With the rapid development of UAV technology, UAV remote sensing has been widely used in agriculture, forestry, resource surveys, and vegetation monitoring [5–10]. Vegetation indices (VIs), as simple and effective measures of the surface vegetation condition, are widely used in vegetation monitoring via remote sensing [11–13]. Because of the unique response characteristics of vegetation in the near-infrared band, most vegetation indices (such as the normalized vegetation index [14] and the soil-adjusted vegetation index) are currently based on a combination of visible light and near-infrared bands [15].

At present, there are a variety of UAV-based multispectral minisensors on the market that can be used for vegetation monitoring [16–23] and can be selected according to the different needs of users. To make the VI products obtained from different sensors at different times comparable, the digital number (DN) of the collected image data is usually converted into reflectance, and then the reflectance is used to calculate the vegetation index [24]. For example, the Parrot Sequoia (Sequoia) multispectral sensor can help users to obtain the reflectance value, and then some conversions can be performed on the reflectance to calculate the VI [25]. Different sensors may use different conversion methods, but the conversion process may have a certain effect on the reflectance value which will further affect the calculated VI value. Unlike the Sequoia, the DJI Phantom 4 Multispectral (P4M) provides users with DN values, and then some conversions can be performed on the DN to calculate the VI. Although these two sensors provide users with different types of data for calculating VI values, the calculation method is the same; i.e., the VI value is the result of the spectral value (Sequoia: reflectance; P4M: DN) after linear or nonlinear transformation. Between these sensors, there is a certain difference between the spectral values in the same band, and this difference may be enlarged or reduced after VI calculations. This leaves the question as to whether there is any difference between the VI products obtained by the above two methods. To answer this question, this study considered two UAV multispectral minisensors, Sequoia and P4M. The research was conducted based on the consistency of spectral values, consistency of VI products, and accuracy of VI products, as these assessment methods have been widely used in different sensors [26–29].

Zhang et al. [30] compared the reflectance and normalized difference vegetation index (NDVI) of the medium spatial resolution satellite Sentinel-2A with those of Landsat-8. The results showed that the Sentinel-2A surface reflectance was greater than the Landsat-8 surface reflectance for all bands except the green, red, and the broad Sentinel-2A near-infrared bands. The Sentinel-2A surface NDVI was greater than the Landsat-8 surface NDVI. Ahmadian et al. [31] estimated the physiological and physical parameters of crops by using the VIs of Landsat8 OLI and Landsat-7 ETM+. The results showed that Landsat-8 OLI was better at capturing small variability in the VIs, making it more suitable for use in the estimation of crop physiological parameters. Roy et al. [32] compared Landsat-8 and Landsat-7 in terms of reflectance and NDVI. The results showed that the reflectance and NDVI of Landsat-8 were both greater than those of Landsat-7. In order to accurately distinguish cassava and sugarcane in images, Phongaksorn et al. [33] compared the reflectance and NDVI of Landsat-5 and THEOS. The results showed that THEOS can better distinguish the two crops. These previous studies show the value of studying different sensor platforms in terms of reflectance and VI in order to evaluate their performance. While there are many other relevant studies focusing on satellite-based sensors [34–38], only a few studies have considered UAV-based minisensors.

Bueren et al. [39] compared four optical UAV-based sensors (RGB camera, near-infrared camera, MCA6 camera, and STS spectroradiometer) to evaluate their suitability for agricultural applications. The STS spectrometer and the multispectral camera MCA6 were found to deliver spectral data that can match the spectral measurements of an Analytical Spectral Devices handheld spectroradiometer (ASD) at ground level when compared over all waypoints. Bareth et al. [40] compared the Cubert UHD185 Firefly and Rikola hyperspectral camera (RHC) to introduce their performance in precision agriculture. The results showed that they both worked well, and the flight campaigns successfully delivered hyperspectral data. Nebiker et al. [41] compared three sensors (Canon s110 NIR, multi-SEPC 4C Prototype, and multi-SEPC 4C commercial) to investigate their characteristics and performance in agronomical research. The investigations showed that the SEPC 4C (multi-SEPC 4C Prototype and multi-SEPC 4C commercial) matched very well with ground-based field spectrometer measurements, while the Canon s110 NIR expressed significant biases. Deng et al. [42] systematically compared the vegetation observation capabilities of MCA and Sequoia based on reflectance and VI. It was found that the reflectance of the MCA camera had higher accuracy in the near-infrared band, and the reflectance accuracy of the Sequoia camera was more stable in each band. The MCA camera can obtain an NDVI product with a higher accuracy after using a more precise nonlinear calibration method.

In recent years, UAV minisensors have begun to show an end-to-end (user to product) development trend, which simplifies the data processing and VI calculation, thus giving users the best sense of use. It is necessary to further explore the performance of different UAV multispectral minisensors based on previous research. To address this, our main objective of this paper was to experimentally evaluate different UAV multispectral minisensors and compare them in terms of consistency. To meet the main objective, we focused on (1) analyzing the consistency of spectral values, (2) analyzing the consistency of VI products, and (3) assessing the accuracy of NDVI products between two sensors. This research will suggest whether vegetation observations from different sensors complement each other or not, thereby further broadening their application in different fields.

#### **2. Materials and Methods**

#### *2.1. Study Area*

The study area is located in Fangshan District, Beijing, China (39◦33 34.93 N, 115◦47 40.97 E), which has a warm temperate humid monsoon climate (Figure 1). It covers an area of 0.03 km<sup>2</sup> with flat terrain and 95 m average altitude. The annual average temperature is 11.6 ◦C, and the annual average precipitation is 602.5 mm. There are a variety of surface types in the area, mainly grassland.

**Figure 1.** Location of study area.
