1. Introduction
The growth of technologies such as 3D scanning and Building Information Modeling (BIM) has been cited as promising in future inspection fields. Additionally, in construction and civil engineering, the use of safety inspections and construction continues to increase [
1,
2,
3,
4,
5,
6,
7].
Laser scanners were developed in 1960, but their application in industrial fields began in the late 1990s. Scanners have also been developed for military purposes, such as drones (UAV), but they are used in various areas. As reported in Ref. [
8], the demand for laser scanners continues to increase in the construction industry, being only 20% in 2016 but increasing to 57% in 2018. The need for scanning technologies is expected to grow annually. Initially, various construction fields were applied using a scanner, such as ex-citation measurements, construction assessments, and construction metrology [
9,
10,
11,
12].
There are a few typical methods that efficiently measure structural deflection [
13]. The most common method, which detects defects and conducts safety inspections, is a visual inspection. However, this method of inspection has limitations in modern complex structures because it is time-consuming, demands trained engineers, and is labor-intensive. This makes it difficult to assess large spatial structures or high-rise buildings. Thus, various studies on remote methods have been conducted to overcome the limitations of contact inspections. With technological development, combined methods based on the conventional and state-of-the-art methods have been used to detect and monitor structural behaviors. A remote detection method using ultrasonic waves and an air-coupled transducer was employed to determine the structural defect using the narrow frequency range of the air-coupled converter [
14]. A way to detect structural defects using photogrammetry was introduced [
15]. Defect detection algorithms that incorporate computer vision and image processing to increase accuracy have been demonstrated. Total robotic stations transfer data even via the Internet and are managed remotely. High precision and measurement automation is characterized [
16]. The GNSS satellite is usually used to survey tall sites and bridges of large spans [
17,
18,
19]. The stereovision-based crack width detection technology using the Canny–Zernike combination algorithm was introduced [
20]. This measured crack width by evaluating the minimum distance between two sides of the crack edge. Compared with the methods that use gauges and vernier calipers for measurements, the accuracy of these methods was higher than that of 2D images.
Notably, various developments such as safety inspections and scan-to-BIM are being made through multidisciplinary converging technologies. The demand for BIM technology is increasing, in line with the development of scanning technology [
21]. BIM is a growing field involving engineers, contractors, and architects. Digital photogrammetry can be relatively inexpensive and highly accurate, and it can offer rapid, remote, and three-dimensional data acquired with images that provide a permanent visual record of the test [
22,
23]. Recently, defects in structural members have been identified using data obtained through 3D laser scanners, and maintenance planning for large-scale facilities has been established [
24]. This research was conducted to estimate the defects of bridges using drones, which install laser scanners and multispectral cameras. Scanners detected debris and cracks by applying loads to the experimental structures [
25,
26]. In addition to modern structures, cultural heritage sites are scanned to establish reverse drawings and safety inspections [
27]. Gordon and Lichti [
28] developed a modeling strategy that allows coarse-precision terrestrial laser scanner observations to accurately measure vertical deflections of deforming beams. Artese and Zinno [
29] proposed a system that uses a terrestrial laser scanner to evaluate the structural health and monitor its bearing capacity, set as a line scanner and positioned under the bridge deck. In addition, several studies have been conducted in Korea to confirm the possibility of technology through field application using several 3D laser scanners [
30,
31,
32,
33].
In this study, the new evaluation system was proposed by reverse engineering using a 3D laser scanner. This system enabled quick safety inspection through 3D point cloud data from scan data. Various applications were discussed to check the possibility of assessment dependent on Korean conditions and the Korean Design Standard (KDS) [
34,
35].
2. Structural Evaluation System of Using 3D Laser Scanner
An evaluation system using a 3D laser scanner is a safety inspection system that quickly and accurately diagnoses the structural information of a building. After acquiring a point cloud based on digital information using LIDAR technology, a precise safety diagnosis of buildings was performed based on the condition evaluation criteria. The goal of the evaluation system using 3D laser scanners is to achieve accuracy, time reduction, and economic feasibility compared with existing safety diagnoses. Trimble R10, RTS-773, and X7 were used in this research and are shown in
Figure 1.
Figure 2 shows the entire process of the evaluation system using a 3D laser scanner. The first step was to acquire structural data. The data acquisition process was divided into targeting and scanning steps. The reference coordinate point is measured using the GNSS equipment and the coordinate point data measured by the total station are entered. Subsequently, the targeting step was completed by measuring the coordinate point of the starting point of the scanning step using the total station. This step can enter the location information (GPS) of the scanned data. This process can increase the objectivity and accuracy of the scan data. Scanning starts at the time of targeting and is performed according to the planned path. The scanner can simultaneously acquire images and point cloud data.
The second step is status assessment. Structure status inspection is conducted using the acquired image. The data are high-definition panorama images that can locate cracks and deteriorated structures through images in the office.
The third step involves 3D modeling using a dedicated program based on point cloud data (PCD). In this study, the PCD editing program, Trimble Real Works, was used. The Real Works program removes noise from the acquired data and sets the necessary parts. Through this work, the high-capacity PCD was reduced to increase the convenience of the work. The necessary parts and objects were classified, and the files were stored for each edited data item based on the edited data. Structural drawings were obtained from these data, and the deformation or deflection was evaluated using these drawings. The final step was the structural analysis step. Based on the data acquired in the previous step, structural analysis was performed using drawings. The current status of the structure is verified through this step.
2.1. Image Data Acquisition
2.1.1. Measurement
Data acquisition using laser scanners is used for targeting and scanning. GNSS and total stations were used as the targets. The data were acquired using a scanner based on the location information obtained during the scanning step. The specifications of the equipment used in this study are listed in
Table 1.
Targeting is a process of inputting GPS data into a PCD obtained through scanning. Two or more actual location positions are received in the 3D laser scanner through the targeting step. When the target point is located underground, GPS data must be obtained using GNSS and the total station from the ground, and the scan path should be connected from underground to outside.
In this study, the scanner measured 500,000 points per second and the measurement range of the data was up to 80 m. Trimble X7 has an automatic calibration capacity that allows horizontal calibration and 10 MP high-resolution image data. Additionally, it is linked to a high-performance tablet, in which the data can be checked in real time. Accordingly, the success of the automatic registration process is matched in real time, and it may or may not require additional scanning. The total station is an electronic instrument used for surveying and building construction. It is an electronic transit theodolite, integrated with electronic distance measurement to measure vertical and horizontal angles and the slope distance from the instrument to a particular point, collect data, and perform triangulation calculations. However, the total station was used to connect the GNSS and scanner in this study. As the 3D scanner cannot enter GPS data directly, GPS data was entered into the total station, and coordinate values were entered into the scanner to start the survey. The coordinate values include GPS data from the total station.
After the scanning was completed, the data were processed. The data acquired two types of results: image data (
Figure 3) and digital PCD data due to initial processing, such as noise removal and section setup in the acquired PCD.
2.1.2. Defect Detection
Visual inspection can be performed through image data acquired by the scanner. The image data are high-definition panorama images that can locate cracks and deteriorated structures through images in the office. For the visual evaluation, several types of detection were checked: crack, exfoliation (breakout), water leakage, efflorescence, etc. The presence of cracks was determined based on the images captured through the laser scanner, but it cannot check the volumetric features of the crack, such as crack depth. The multidisciplinary photogrammetry techniques with several innovative techniques should be used to inspect in detail. For example, in the case of the evaluation for surface aging, it can be performed through state evaluation by measuring the area of the defect.
2.2. Deflection and Deformation Detection
2.2.1. Post-Processing
Initial processing is the step of refining data so that high-capacity PCD can be handled more conveniently. The point cloud methodology reconstructs the 3D surface of an object through an alignment of extracting and matching feature points of an object photographed from various angles and represents the point cloud as a 3D surface [
36]. In this paper, the laser measures the distance with the time-of-flight and enters the value into the device. The point cloud is configured by shooting a laser. The 3D location information of the point cloud is configured based on the obtained GPS location coordinates.
Post-processing is the task of classifying data according to their purpose. Post-processing automatically classifies and organizes data for each object using an automatic classification technique (Auto-Classify (Indoor)) among ‘Real Works’. The technique automatically classifies the objects, such as structural members, men, vehicles, etc. Sometimes, it is not possible to classify structural members, such as a small object located in a corner, similar in color to other structural member. Under this condition, classification is performed manually. Subsequently, a structural evaluation is conducted, focusing on the data of the structure. Therefore, the data classified as non-structural members can be classified. In addition, the scanner location data are represented by a triangle within the program. Through this, the measurement location can be confirmed (
Figure 4).
2.2.2. Detection Based on Scan Data
The 3D laser scanner measures the distance by the time the shot laser hits an object and returns. It is scanned from various angles, and then matching is conducted based on overlapping points. Through this process, a 3D model is constructed in a point cloud. In this study, a 3D PCD model was created using the Real Work commercial software. Moreover, a structural drawing was created based on the location information (GPS) data obtained during the targeting step. Structural drawings created with reverse engineering can be used to detect building tilt and vertical displacement measurements through comparative analysis with existing drawings (
Figure 5). Inspecting aging buildings without drawings can be performed faster and more accurately than conventional methods using reverse engineering.
The 3D point cloud model can measure deflection and deformation, such as tilting and vertical displacements. Point clouds were generated from the GPS data and represented in a 3D coordinate space with the transformed GPS data. With the locations developed, it was possible to measure the deflection of the slab and column deformation. The edges were calculated as vectors, and slopes were measured as the angles between the sides and floor. Vertical and horizontal displacements of the edge of the building were estimated through comparisons with 3D coordinate location data of each floor points. The 3D model, generated using the laser scanner, uses GPS to calculate the position coordinate of each point of the point cloud, which can be used for measuring the selection and deformation. It is possible to calculate the height and lateral displacement and check the value of the election and deformation given in Equation (1) (
Figure 6) [
37,
38,
39].
2.3. FEM Structural Analysis
A structural analysis was conducted based on structural drawings from reverse engineering to evaluate the safety of the structure. In this study, finite element analysis (FEM) was performed using the CSI SAP 2000 commercial software. Through finite element analysis (FEM), the impact on the structure was analyzed by applying loads. Based on previous studies, FEM structural analysis of the floor was carried out [
40,
41,
42,
43,
44,
45,
46,
47,
48,
49]. Furthermore, a structural analysis was conducted based on existing drawings to compare the two analysis results. The amount of deformation that progressed on the structure could be detected by comparing and analyzing the results of the two analyses.