Methods of mapping the lands disturbed by mining operations and accuracy of cartographic images obtained from Unmanned Aerial Vehicles : A review

Purpose. Analyzing the land disturbance consequences caused by surface mining operations and methods for mapping these lands, as well as studying the accuracy of point coordinates of digital images obtained from materials of aerial photographic surveys using Unmanned Aerial Vehicles (UAVs). Performing a quantitative assessment of the Root Mean Square Error (RMSE) of point coordinates on cartographic images and determining the dependences of the RMSE of point coordinates on the photogrammetric parameters. Methods. The review of previous research publications within the framework of the presented subject is performed in the following sequence: analysis of ecosystem disbalance as a result of surface mining operations; based on previous studies, collecting the data for quantitative assessment of accuracy in the form of RMSE of point coordinates on cartographic images obtained from the materials of aerial photographic survey using UAVs; statistical study of the relationship between the RMSE and photographic survey parameters. Findings. The methods for mapping the disturbed lands to return them to their natural state after the consequences of surface mining operations are presented, based on a review of previous research publications on the subject of the work. According to the previous studies, the RMSE of point coordinates of cartographic images has been systematized, and, based on this, the accuracy of topographic plans has been determined for them. Statistical studies of the relationship between the quantitative assessment of the RMSE (xy) and RMSE (z) accuracy in relation to the photographic survey parameters have been performed. In addition, the scattering diagrams of the correlation dependence and the range of RMSE relative frequency have been presented. Originality. Based on a critical analysis of previous studies on the lack of quantitative accuracy regulation of cartographic images obtained from aerial photographic survey using UAVs, the RMSE systematics has been performed in terms of the photographic survey height. Based on this, the accuracy of topographic plans, the relative frequency of horizontal and vertical distribution of errors, the mean value x and the root mean square error (σ) have been determined. Practical implications. The systematics of the RMSE values of cartographic image point coordinates for certain photographic survey parameters and the scale of topographic images makes it possible to take this into account in the project of aerial photographic survey using UAVs of lands for various purposes, as well as to choose the height and photographic equipment according to the required accuracy.


Introduction
Mining of mineral deposits occupy a significant place on the territory of Ukraine, but surface amber mining has become the most popular. In the forests of Rivne, Zhytomyr and Volyn regions, due to illegal mining of amber on forestry lands, the area of coniferous forests is decreasing and the area of disturbed lands in need of reclamation is increasing [1]- [3].
Surface mining of amber on forestry fund lands has led to irreparable consequences. The vegetation cover has lost the integrity of the grass cover, as well as the destruction of tree plantations and the drying out of the forest stand are taking place [1], [5]. The damage from illegal mining has affected the environment, leading to the degradation of the existing relief structure, pollution of underground and surface water resources, etc. As a result of surface mining, small artificial lakes emerge in the form of openings, where water pumps directly flood the holes that facilitate the formation of small lakes [6], [7]. Turkish scientists [8] have developed a methodology to determine the prospects for land use of disturbed areas to reduce environmental damage, using the example of marble mining in the regions of Antalya, Burdur and Isparta. The influence of tin mining on changes in the environment and natural landscape on the Belitung Island in Indonesia has been studied. Therefore, all disturbed lands, as well as areas adjacent to them, which have completely or partially lost productivity, caused by surface mining, are subject to recla-mation. The latter is important for ensuring and returning these lands to their natural state [9]- [11].
Taking into account the ecological catastrophe in the areas of amber mining, the Cabinet of Ministers of Ukraine in 2017 adopted a resolution approving the "Procedure for the implementation of a pilot project for reclamation of forestry lands disturbed by illegal amber mining", using the example of forest areas of the No. 59 division of Dubrovytsia Forestry in Dubrovytsia District of Rivne Oblast. According to the data of the report "On environmental impact assessment of reclamation of lands disturbed by illegal amber mining", the total area of disturbed forestry lands in the No. 59 division of Dubrovytsia Forestry is 71 hectares, and the total area of disturbed plots is 69.9 hectares, that is, 98% of the division area (Fig. 1). A significant part of the forestry land area is disturbed by the use of motor pumps in pits and hydrodynamic bell-pits during illegal mining of amber, which leads to the destruction of the natural landscape [12].

Figure 1. The location of land plots disturbed by amber mining within the No. 59 division
At the stage of planning the reclamation works, it is important to determine the expediency of further use of these land areas and substantiate the direction of reclamation [13]. Based on the example of a mining site in Liaoning Province, China, a reclamation strategy has been developed that includes the land suitability analysis and assessment of ecosystem services. Assessment of ecosystem services helps to make decisions and implement forestry methods for reclamation of forest areas after technogenic processes caused by human activities [14].
For the design of engineering structures for reclamation systems and reclamation projects, it is necessary to have high-quality topographic and cartographic material that would make it possible to draw up rational projects for planning the disturbed lands with the required accuracy. Today, cartographic images in the form of topographic maps and plans are obtained using modern methods of the Earth's Remote Sensing (ERS), in particular, from Unmanned Aerial Vehicles (UAVs), which are an alternative to traditional methods. In recent years, satellite survey method has been used to map lands and study environmental changes caused by surface mining. The satellite survey method is important for monitoring the lands disturbed by illegal mining, as well as identifying and tracking of reclamation sites, and assessing changes in soil cover [15]. Monitoring for the dis-turbed lands within the No. 59 division of Dubrovytsia Forestry was performed using multi-temporal satellite survey method of the Landsat-5TM, Sentinel-2A and Sentinel-2B series of 2009, 2015 and 2017. As a result, the forest areas to be recultivated in the first place, and which areas should be left for natural restoration, have been determined. The restrictions in the use of satellite survey method for land mapping are the spatial resolution of materials, the cost and time required to process them, and weather conditions, including thick clouds. According to aerial photographic survey data from UAVs, digital elevation models (DEM) with high spatial resolution can be obtained, requiring field inspection to obtain high-precision topographic materials [16], [17].
Digital terrain mapping is a complex of processes for collecting and processing digital topographic information, forming a digital elevation model in software, which makes it possible to obtain various analytical and graphical materials. Digital surface models (DSM) are created in such a way that it is possible to select from them a model of terrain relief, building, structure, communication, hydrography, vegetation cover. Digital information about the terrain can be obtained in different ways: ground based geodetic method using electronic tacheometers and Global Positioning System (GPS); cartographic method, for which ordinary topographic maps serve as digital information about the terrain. The advent of small and inexpensive UAVs for aerial photographic survey has contributed to the development of new methods and technologies of digital mapping. Aerial photographic survey using UAVs has an advantage over satellite survey method; based on satellite survey data, the maps have an overall accuracy of 78.1%, and the accuracy of maps based on UAV data is 92.3%. The data obtained from UAV provide better detail of topographic images, including relief elements (hills, slopes, ravines, landslides, orographic lines), for land management [18]. Aerial photographic survey using UAV does not fundamentally differ from photography from manned aircraft, but it has its own characteristics in terms of its cruising speed, range of flight altitudes, conducting and installing main and additional equipment on UAV board. Land mapping using UAV is possible only if the geometric parameters of traditional aerial photographic survey are observed. The accuracy of such mapping can increase tens of times and be about a pixel (ground sample distance -GSD) for both conventional aerial photographic survey and satellite survey method [19]. It is known from traditional photogrammetry that the accuracy of aerial photographs depends on the photographing altitude, camera calibration and the accuracy of the control point coordinates connected with the ground geodetic network. The cost of mapping using traditional aerial photogrammetry is very high. Thanks to the introduction of small-format aerial photogrammetry technology, it became possible to obtain a digital map from digital aerial photographs of a small-format camera mounted on a lightweight UAV platform [20], [21]. Aerial photographic survey using UAVs has found its use for mapping the forest lands, monitoring and controlling the expansion of violations from illegal mining, forestry management, etc. Remote sensing from UAVs in forestry is still in the experimental stage, but it is expected to expand rapidly [22].
Over the last decade, a significant amount of research has been performed on the accuracy of mapping and 3-D modeling based on UAV data. The accuracy and substantiation of mapping materials is the main requirement for any new method of constructing topographic plans, digital surface models (DSM) and digital elevation models (DEM), including aerial photographic survey using UAV [23]. Remote sensing materials obtained from UAV are important for monitoring, including calculating the volume of soil displaced during construction and surface mining, studying the patterns of vegetation development, as well as observing the topographic changes, including subsidence, erosion and the development of streams [24]. The restriction in the Earth's remote sensing takes for the accurate identification of small areas and linear objects, the spatial resolution of a raster grid, which must be many times more accurate than an object in order to dtermine its peculiarities in detail. Accurate largescale mapping of small, fragmented and linear vegetation areas depends on their area and extent [25]. Aerial photographic survey using UAV is used to create digital topographic plans of settlements at a scale of 1:2000, therefore, new ways to improve the use of small UAV systems for aerial photographic survey have been identified in [26]. Recently, digital elevation models (DEM) and orthophotomaps have been supplied in digital format and used in combination as an alternative to traditional topographic and contour maps. Large-scale topographic plans and DEMs are used to draw up projects for drainage, irrigation, land reclamation for agricultural and forestry purposes, in the mining and oil industries, as well as for predicting the hazards in the ecological system, which makes their accurate creation very important [27]- [29].
Land mapping using UAV makes it possible to obtain high-quality digital cartographic images, if to observe certain requirements for the process of implementation and setting photogrammetric equipment for aerial photographic survey, as well as the image processing algorithm. Aerial photographic survey using UAV provides spatial and temporal resolution for mapping and monitoring of natural landscapes and lands for various purposes. Modern specialized software makes it possible to automatically obtain DSM and DEM in the form of a dense 3D cloud of digital points based on aerial photographic survey data using UAV. Recently, studies on assessing the accuracy of cartographic images obtained from UAV data according to the number and distribution of ground control points (GCP), obtained in the course of work (VAU-RRK-SfM), have become widespread. Processing of aerial photographic survey materials obtained from UAV using Structure from Motion (SfM) software has become widespread in different countries around the world. The application of VAU-RRK-SfM makes it possible to obtain point clouds for a 3D-model based on high-resolution aerial photographs from UAV. The georeference of the point cloud to the GCP can achieve an accuracy of up to 25-40 mm based on images obtained from a flight altitude of approximately 50 m [30]- [33]. Therefore, obtaining the coordinates of ground control points remains today a very important, but time-consuming task. Conventional georeference using GCPs provides reliable positioning, but geometric accuracy is critically dependent on the number and distribution of GCPs. Direct georeferencing of images from UAV using a GNSS differential correction such as PPK (post-processing kinematics), overcomes these restrictions by providing accurate and directly georeferenced mappings [34]. It is indicated in the previous studies that with the use of modern equipment and GNSS geodetic sensor on UAV board, pixel accuracy of up to 3-5 cm can be achieved without GCPs [35], [36]. The issue of using the ground control points (GCP) for georeferencing of cartographic images is still relevant today. According to a review of publications, many previous studies are devoted to this problem with different approaches to its solution [37], [38].
Modern software for digital photogrammetry is capable of creating digital orthophotomaps and digital maps for both large-format metric camera and small-format digital camera. As for the accuracy assessment, it cannot be denied that the accuracy for a large-format metric camera is higher. During the review of previous studies, it has been substantiated that high accuracy can be achieved with a large-format metric camera compared to the accuracy of a small-format digital camera. Today, using a small-format digital camera, submeter accuracy can be achieved [21]. In recent years, as can be seen from the review of publications, previous studies have shown the validity of assessing the accuracy of cartographic images obtained from UAV data. From the analysis of previous studies, it has been found that the quantitative accuracy is somewhat different, and the accuracy of each individual research differs from the other. In the presence of a significant number of the RMSE of point coordinates and photogrammetric parameters, it is possible to reveal the actual RMSE range of quantitative accuracy of the cartographic image point coordinates. The lack of analysis and systematization of the disparate quantitative RMSEs of point coordinates is the motivation for our research.
Thus, the purpose of this paper is to review methods for mapping the lands disturbed by mining operations for their further reclamation and to systematize the RMSEs of point positions of the cartographic images obtained from UAV materials, taking into account the influencing factors, and also to determine the correlation between them based on an analytical review of previous research publications on this subject.
To achieve this purpose it is necessary to solve the following tasks: to review the methods of mapping the lands disturbed by mining operations; to systematize quantitative RMSEs of point coordinates of cartographic images obtained from UAV materials and determine the scale accuracy of topographic plans; performing a statistical analysis of the ratio between the RMSE accuracy of cartographic image points and photographic survey parameters.
Today this task is relevant and requires further research.

Materials and methods
When performing the research, the authors use regulatory legal documents of the state authorities of Ukraine, the materials of the report "Environmental Impact Assessment" made by the Institute of Environmental Geochemistry of the National Academy of Sciences of Ukraine, a review of research publications on the methods for accurately mapping of lands disturbed by surface mining of domestic and foreign authors, special literature, etc. This research is performed in the following sequence: analysis of the accuracy of cartographic images obtained from aerial photographic survey materials using UAV; an overview of the factors influencing the accuracy of digital cartographic images; analysis of root mean square errors of cartographic image points and their systematization by flight altitude.

Analysis of the accuracy of cartographic images obtained from aerial photographic survey materials using UAV
One of the criteria for assessing the accuracy of creating a digital elevation model based on UAV results is the RMSE of determining the point coordinates of cartographic images or the mean deviation of the referencing point coordinates in the triple overlap area of aerial photographs. In classical photogrammetry, the accuracy of the digital orthophotomap mofp is characterized by the error in measuring the position of the contour points on the orthophotomap: where: Msn and Mofpdenominator of the scales of the original photographs and orthophotoplan; mΔrerror of orthophotomap images, conditioned by the influence of the terrain relief; mf and mrerrors of photogrammetric and geodetic measurements.
The permissible accuracy of the relief influence mΔr dop on the DEM is determined by the measurement errors and the permissible orthophotomap error mofp dop [39].
According to the Instruction for Topographic Survey at scales of 1:5000, 1:2000, 1:1000 and 1:500, "the maximum error of planned survey network points (including identification marks) relative to points of the state geodetic network and extensive geodetic networks should not exceed 0.2 mm at a scale of the plan in open and in the built-up areas, and 0.3 mm in the area covered with trees and shrubs" [40].
The mapping accuracy of aerial photographic survey using UAV depends on the RMSE accuracy of coordinates of ground control points, their number and distribution [41], [42]. For the most part, the RMSE quantitative accuracy is calculated as the difference between the coordinates of ground control points obtained from processing the images and the coordinates of these points from field measurements. It should be noted that in order to obtain an accurate orthophotomap and DSM, it is necessary to see this model stereoscopically, control the digital elevation model (DEM) using selected points for constructing the relief and performing metric measurements in absolute project coordinates, etc. At the present stage, stereoscopic observation of photographs from UAV is implemented in the software algorithms of digital photogrammetic stations (DPS). For example, the software for processing the images obtained from UAV is implemented in PHOTOMOD UAS by Rakurs Company, which makes it possible to create correct orthophotomaps for cartographic purposes [42].
The accuracy of cartographic images improves significantly depending on GSD (ground sample distance) resolution. Since the GSD (m/pixels) of images is calculated relative to the flight altitude and camera resolution, the vertical error increases as the GSD decreases [42]. The spatial resolution (GSD) of the image is higher for a lower altitude and, vice versa, it is lower for a higher flight altitude. The resolution from the ground surface of 24 and 50 mm can be achieved from the UAV flight altitude of approximately 80 and 160 m. Based on the previous studies, it has been revealed that the resolution of digital elevation model (DEM) is determined not only by the flight altitude, but also by the software used to process the research data. For example, for processing data obtained from UAV, the PhotoScan software package for computer vision with a height resolution of 10-17 mm is the most effective in flat terrain, while the photogrammetric Rapid Terrain with a resolution of 30-40 mm has advantages in high-relief terrain [33], [43].

An overview of factors influencing the accuracy of digital cartographic images
In a review of most previous research publications, it is noted that the accuracy of digital surface model (DSM) and orthophotoplan mainly depends on the quality and stability of photogrammetric equipment, photographing altitude, image scale, image overlap percentage, number and distribution of GCP, as well as processing algorithm. Based on a review of previous research publications, a significant influence on the accuracy of the constructed DSM has been noted not only by such factors as the photographing altitude, the number of planned and high-altitude ground control points, but also by the accuracy of determining the GCP coordinates. The accuracy of topographic images, orthophotoplan and digital surface model depends on the number of GCP, their distribution over the area and the configuration of the studied object [44]- [48]. In separate publications, the relationship between the errors of digital surface model points, their horizontal distance to the nearest GCP and flight altitude has been studied [42]. Although sometimes a higher flight altitude gives better accuracy based on other factors: the presence and number of GCPs, a geodetic receiver on the UAV platform, the percentage of image overlap, weather conditions and software for data processing [42], [49]. However, increasing the image overlap percentage has a positive influence on DSM accuracy, but leads to an increase in the number of images and the time to process them. Therefore, the image overlap percentage and scope for processing during an aerial photographic survey project is a trade-off between DSM accuracy and time spent for processing [48], [50]. The difference in time and cost of processing the aerial photographic survey projects with minimum and maximum image overlap percentages does not make a big difference in the survey data [28]. In order to reduce the number of images by increasing the UAV flight altitude, processing time and not degrading the DSM accuracy, a method for the radiometric transformation of images obtained using UAV at a high flight altitude (500 m) has been studied and compared with photogrammetric materials obtained from UAV at a low flight altitude. Such a new approach to studying the influence of radiometric transformation on the images obtained from UAV at a relatively high altitude can overcome the problems with low flight altitude [49], [50].

Analysis of root mean square errors of cartographic image points and their systematization by flight altitude
The purpose of this paper is to provide a review of the previous research publications on the accuracy of cartographic images obtained using UAV in the form of RMSE of point coordinates. The RMSE of point coordinates is used as a criterion for assessing the accuracy of cartographic images obtained using UAV.
Based on the previous studies, it has been revealed that most often the analysis of digital surface model (DSM) accuracy is performed by ground control points, which are recognized on the orthophotoplan to obtain Х and Y coordinates, and the Z height is found by DSM. In some studies, RMSE of point coordinates is determined as the difference between the coordinates of ground control points measured on the orthophotoplan and those measured in-situ using the ground method, or as the difference between the referencing points of the photogrammetric network on different images. As previous studies have shown, the difference in the control point coordinates obtained by traditional ground methods and using aerial photographic survey from UAV does not exceed 1-3 cm in plan and 10 cm in height [42]. The research results [46] have confirmed that the horizontal accuracy of RMSE xy of point coordinates is not depend on the flight altitude and terrain relief, and their accuracy increases with an increase in the number of GCPs. The vertical accuracy of RMSE z of a point is not depend on the relief, but the flight altitude and the number of ground high-altitude GCPs have a significant influence [30], [34], [48]. According to the research results [42], the value of the vertical accuracy of the digital elevation model (DEM) for flight altitude 300 m is 0.11 m, and for altitude 500 m -0.15 m. Similar image accuracies have been obtained for high and low flight altitude due to radiometric transformations [49], [50]. According to the Instruction for Topographic Survey, the mean errors in the position of GCPs (taking into account the recognition error on the images) should not exceed 0.1 mm at a scale of the developed map, and the maximum error should be 0.2 mm. The maximum errors in determining the coordinates of the planned GCPs in-situ should not exceed 0.14 mm on the map scale, and in areas covered with forests or bushes, the maximum error is 0.30 mm. The mean errors in determining the heights of GCPs (taking into account the recognition error on the images) should not exceed 0.1 of the contour interval height, and the maximum error should not exceed 0.2 of the contour interval height.
A review of previous research publications on quantitative accuracy has shown that the accuracy of the point positions for the same flight altitude is different, and the accuracy of an individual research can differ significantly from the other. Thus, for example, RMSE in studies [21] for a flight altitude of 100 m in plan is 0.28 m, and for a flight altitude of 0.32 m, in studies [41], it is 0.02 and 0.04 m, respectively. The discrepancies are 0.262 in plan and 0.284 m in height, which do not meet the requirements for photogrammetric works. In order to substantiate the reliable values of cartographic images obtained from UAV, it is necessary to have a significant number of samples for assessing the accuracy of previous studies. Taking into account the indicated previous research publications, in particular [22], [34], the RMSE have been supplemented with preliminary studies [51]- [55].

Results and discussion
In order to determine the actual accuracy of point coordinates on cartographic images according to the scale, based on preliminary studies, quantitative RMSE of point coordinates and photographic survey parameters have been collected, which are presented in Table 1. The quantitative RMSE values and the flight altitude of previous studies have been grouped to determine correlation dependence between them. The value of the approximation authenticity (R 2 ) and the correlation coefficient (r) are taken as indicators of statistical relationship. The statistical studies performed according to data in Table 1 have revealed that the correlation ratios between the quantitative horizontal RMSE (xy) and vertical RMSE (z) of point coordinates and flight altitude are rx,y = 0.86 and rz = 0.62, respectively. The correlation dependence of vertical RMSE (z) is less, which can be related to the number and accuracy of in-situ control points for highaltitude georeferencing of the network (Fig. 4а, b).   Figure 5а, b shows that there is no a correlation ratio between the quantitative accuracy of horizontal RMSE (ху) and vertical RMSE (z) and resolution (GSD), as evidenced by (R 2 ) correlation coefficients of rxy = 0.44 and rz = 0.32. This means that the accuracy of point coordinates is dependent on more significant factors, such as: the accuracy of determining the GCP coordinates in-situ and the image processing algorithm. There is a closer ratio between GSD and flight altitude, the correlation coefficient is r(GSDA) = 0.59 (Fig. 5с), which is confirmed by a review of previous research publications. The scattering diagram of the vertical accuracy of RMSE (z) relative to the number of GCPs shows that the results of previous studies do not show closeness of the ratio (Fig. 5d). Figure 6 presents a histogram of the relative frequency of the vertical and horizontal distribution of RMSE obtained from previous studies of the quantitative accuracy of cartographic images. According to statistical analysis, the vertical distribution of RMSE (z) for most of the previous studies ranges from 0.02 to 0.07 m, the mean value of errors is 0.07 m, and the standard deviation (σ) of these vertical errors is σz = 0.05 m. The horizontal distribution of RMSE (ху) ranges from 0.01 to 0.10 m, but the mean value is 0.07 m and the standard deviation is σxy = 0.07 m. In both cases, the mean values are 0.07, and the standard deviations differ σz = 0.05, and σxy = 0.07 m, that is, they are different. The values of these indicators have been influenced by different instability of RMSE in the sample. * For the overall analysis, these RMSEs are not taken into account because their discrepancy with the others is out of tolerance limit

Figure 5. Resolution scattering diagram of GSD (m/px): relative to accuracy of RMSE (ху) and RMSE (z) errors (а), (b); relative to flight altitude (c); scattering of vertical RMSE (z) error relative to the number of GCPs (d) based on the results of previous studies
In addition, the quantitative RMSEs of point coordinates relative to the UAV flight altitude have been systematized, taking into account the requirements of the Instruction on the accuracy of positions of cartographic image points (Table 2).

Conclusions
Based on a review of previous research publications, it has been noted that the mapping of lands disturbed by surface mining for their further land use is a topical issue for most countries of the world. These publications analyze the methods of land mapping, aerial photographic survey and the accuracy of digital cartographic materials obtained from UAV data.
It has been determined that most of the previous studies have focused on aerial photographic survey projects using UAV, photogrammetric equipment, and image processing software. At the same time, there is a discussion in studies on the availability of GCP. In some studies, it is recommended to use GCP to determine the accuracy of cartographic images, in others, on the contrary, similar accuracy is achieved without GCPs. In our opinion, this problem still requires a thorough statistical analysis.
A review of previous research publications has shown that quantitative assessments of accuracy vary slightly, and the accuracy of each individual research differs from the other. To determine a reliable assessment of the accuracy of point coordinates from previous studies, a significant number (42) of the RMSE of point coordinates and photogrammetric parameters have been collected. Based on this, the RMSEs of point coordinates by the flight altitude have been systematized. To do this, according to the relative frequency statistical histograms, the RMSE values have been found, which are the most common in the sample and taken as the actual RMSE. On this basis, the accuracy of the topographic plan scale and the contour interval height has been determined. The mean values of x and standard deviations σ show the constancy or different variability of RMSEs in the samples. Systematization of RMSEs according to statistical analysis has revealed that the best flight altitude of UAV is 150 and 200 m, for which the root mean square errors are 0.05-0.09 m. We have confirmed the statistical dependence of RMSE accuracy and resolution on the UAV flight altitude, but it has not been revealed between RMSE of point coordinates and resolution.
For mapping the forest lands and lands disturbed by surface mining, for their further reclamation, decimeter-scale accuracy can be taken and, therefore, there is no need to use expensive geodetic survey methods, aerial or ground-based laser scanning, and project ground control points.
It is known that the accuracy of topographic plans, digital surface models (DSM) and orthophotomaps depends on the accuracy of ground point coordinates of the planned-highaltitude geodetic network, which is the basis for the development of phototriangulation of an aerial photographic survey project. However, in view of previous studies, the accuracy of determining the coordinates of ground control points is not classified as an influencing factor, which should be taken into account in subsequent studies.
Given the variety of information obtained from land mapping using UAV and the varying accuracy reported in previous studies, further research is needed to assess the accuracy of materials obtained from UAV in geodetic programs.