Aerial photography
Alireza Afary
Abstract
Extended Abstract
Introduction
3D similar transformation is used in various applications such as photogrammetry, geodesy, robotics and machine vision. Calculating the parameters of this transformation using the least squares method requires determining the initial values as close as to the final ...
Read More
Extended Abstract
Introduction
3D similar transformation is used in various applications such as photogrammetry, geodesy, robotics and machine vision. Calculating the parameters of this transformation using the least squares method requires determining the initial values as close as to the final values. If the initial values used are not close to the final values, especially in the case that the rotation angles related to this transformation have large values, the least squares method will either not converge or converge to a wrong solution. In this paper, a direct and new closed-form method for determining the parameters of this transformation is presented. This method is able to determine 3D similar transmission parameters by using at least three corresponding points in both model and ground coordinate systems. In general, direct and non-iterative methods are faster and have lower computational cost, and most importantly, they do not require initial values. In contrast to these advantages, these methods are sensitive to noise in observations and outliers and have less accuracy than iterative methods. Iterative methods, although they have better accuracy, on the other hand, have more computational cost and their speed is low. Most importantly, these methods require initial values and if the initial values used in these methods are not close enough to the final values of the parameters, these methods will either not converge to the correct solution or converge to a wrong solution.
Methods and Materials
The method presented in this article is based on one of the characteristics of 3D similar transformation, i.e., establishing the same 3D similar transformation relationship between the gravity centers of corresponding points. By transferring the origin of the coordinate systems of the corresponding points to the gravity center points, the 3D similar transformation parameters between these two sets of points can be calculated in a closed-form manner, with the presented method. Two datasets were used to show the effectiveness of the presented method. The first dataset was created by simulation with large rotation angles and four times scale factor and with the minimum number of required points, i.e., three points. To simulate the real state in this dataset, random errors with normal distribution were added to each set of the corresponding points. The second dataset was selected from the real data obtained from LiDAR operations.
Results and discussion
The results of the method presented in this article were compared and evaluated with the results of the least squares method and two other closed-form and direct methods, i.e., the SVD method and the dual quaternion method. The results of the method presented in this article are close to the final values of these parameters and the values obtained from other methods. Tables (6) and (8), respectively, show the difference values of 3D similar transmission parameters between the results of using direct and closed-form methods with the least squares method for simulated dataset and real LiDAR dataset.
The data in Tables (5) and (8) show that the presented closed-form method in this paper provides similar 3D transmission parameters for both simulated data sets and real data with a slight difference of about 0.02° for rotational parameters and with a slight difference of less than 0.2m in the displacement vector parameters and with a slight difference of less than 0.002 in the scale parameter.
Conclusions
As can be seen from the obtained results, the accuracy of the values calculated by the presented method in this article is to the extent that it can be used directly for most applications, especially in online applications. On the other hand, the lower volume of calculations of the method presented in this article, compared to the SVD and dual quaternion methods as well as the iterative least squares method, justifies the use of this method for online applications. Also, the results of this method can be used as accurate initial values for the least squares method, in Close-range and UAV photogrammetry applications, where the rotational angular parameters can have large values.
Aerial photography
Morteza Heidarimozaffar; Reza Zerafatyjamal; Hossein Torabzadeh Khorasani
Abstract
Extended AbstractIntroduction:The production of topographic maps by drone photogrammetry has replaced traditional mapping in many civil projects. UAV photogrammetry due to the rapid development of hardware and software can be used in many geomatic applications including agriculture, forestry, archeology ...
Read More
Extended AbstractIntroduction:The production of topographic maps by drone photogrammetry has replaced traditional mapping in many civil projects. UAV photogrammetry due to the rapid development of hardware and software can be used in many geomatic applications including agriculture, forestry, archeology and architecture, emergency management, and traffic monitoring. Nowadays, mapping the boundaries of linear projects such as roads, water transmission canals, power transmission lines (electricity and gas), railway lines, and the like, is done by UAV photogrammetry. UAV photogrammetry has emerged as a viable alternative due to its lower cost, greater spatial and temporal resolution, and flexibility in capturing images compared to other conventional aerial and space methods. Despite various researches in this field, the effect of the imaging method on the accuracy of topographic maps has not been directly studied and tested. In this paper, the experimental conditions of UAV photogrammetric imaging methods in two modes of height-fixed and scale-fixed were considered. The effect of using accurate positioning equipment on image centers on reducing the number of required control points has also been investigated.Materials & Methods:Assessing and ensuring accuracy is very important for topographic maps. Knowledge of the number and distribution of control points throughout the project area and the optimal imaging method are effective in producing these maps. In this paper, to achieve the most optimal imaging mode and the highest accuracy in the production of topographic maps for this type of project, the effects of flight design parameters, number, and distribution of control points were studied. Also, the effect of using accurate positioning equipment of image centers on reducing the number of required control points has been investigated. Image processing without accurate information of image centers with control points, image processing with accurate information of image centers and without using control points, and image processing with accurate information of image centers and control points, were considered in two flight modes with fixed altitude and fixed scale. Of the 25 points whose exact coordinates were measured differentially by the Global Positioning System, 18 were ground control points and 7 checked points. The control of the elevation coordinates of the points was performed using the direct geometric leveling method. To evaluate the accuracy of the results, the amount of error and the accuracy of the work performed, the leveling operation was performed in a round trip between all control and checkpoints.Results & Discussion:Evaluating the accuracy of the UAV photogrammetric method in producing topographic maps related to this project, according to the criterion, the root means square of errors has been done for the control and checkpoints. In addition to calculating this criterion for control and checkpoints, the difference between the digital models prepared and the reference model was considered another criterion for comparison and evaluation. According to the results, the mean height error of all points in the constant-scale mode has the lowest value in Scenario 3. The numerical value of the mean error, in this case, was equal to 0.010 meters for control points and 0.020 meters for checkpoints. The accuracy of the models obtained from point clouds with dimensions of 0.5 meters is higher than the point clouds of 2 and 4 meters. The largest difference from the reference model is related to model 1 in the category of models obtained from 0.5-meter point clouds, which vary numerically in the range (-1.68 to 10.56) meters.Conclusion:The results of the height error evaluation of control and checkpoints show if aerial triangulation and justification of images use image center observations alone, in challenging projects the mapping of linear areas where the mapping is done in a strip area will not have the desired accuracy. The use of ground reference images alone is not sufficient and the simultaneous use of ground control points and ground reference images improves the accuracy of the results. As a result, the use of ground control points, fixed scale images, and image center information in the image processing process to produce corridor maps provides the best elevation accuracy compared to other modes. Also, the use of the initial digital elevation model of the project area in performing flight operations and capturing images with fixed scales has a significant effect on increasing the accuracy of the elevation component. Based on the comparison of the final digital elevation models compared to the reference model, the accuracy of the models obtained from the resolution of 0.5 meters is higher than 2 and 4 meters. Also, the effectiveness of filters to correct and reduce errors in digital elevation models has better results.
Aerial photography
Zahra Azizi; Mojdeh Miraki
Abstract
Extended Abstract
Introduction
Advances in computer vision and the development of remote sensing instruments have made indirect measurement of tree features possible. Individual tree crown delineation is an important step towards information collection and mapping trees in an urban area. This information ...
Read More
Extended Abstract
Introduction
Advances in computer vision and the development of remote sensing instruments have made indirect measurement of tree features possible. Individual tree crown delineation is an important step towards information collection and mapping trees in an urban area. This information is then used to help planners design strategies for optimization of urban ecosystem services and adapt to climate changes. Common methods of Individual tree crown delineation (ITCD) were based on very high-resolution satellite or Light Detection and Ranging (LiDAR) data. However, satellite data are usually covered by clouds and thus, cannot be appropriate for the measurement of individual trees. Aerial Laser scanning is also relatively expensive. Remote sensing with unmanned aerial vehicle (UAV) captures low altitude imagery and thus, is potentially capable of mapping complex urban vegetation. Automatic delineation of trees with UAV data makes collection of detailed information from trees in large geographic and urban regions possible. Therefore, a multirotor UAV equipped with a high-resolution RGB camera was used in the present study to obtain aerial images and delineate individual trees.
Materials & Methods
The present study has compared the performance of Inverse watershed segmentation (IWS) and region growing (RG) algorithms using point clouds derived from Structure from Motion (SfM) algorithm and UAV imagery captured with the aim of tree delineation in Fateh urban forest located in Karaj. Region growing (RG) is used to separate regions and distinguish objects in an image. It starts at the initial seed points and determines whether the neighboring pixels should be added to the growing region. If the neighboring pixels are sufficiently similar to the seed pixel, they are labeled as belonging to the seed pixel. To implement the algorithm, the window size and the growing threshold were set for all resolutions. In order to obtain the most appropriate size for the search window, we examined different window sizes with a growing threshold of 0.5 for each resolution. Individual trees delineation was performed for each CHM resolution in the three different sites using "itcSegment" package of R software. Watershed segmentation algorithm is also similar to RG algorithm. The only difference is that it sets the growing seeds at the local minima. In other words, the local maxima in this algorithm change into local minima and vice versa. Inverse Watershed Segmentation (IWS) method was implemented in ArcGIS 10.3 because of its capability in delineation of distinct tree entities. In the summer of 2018, three sites with different structures including a mixed uneven-aged dense stand (site 1), a mixed uneven-aged sparse stand (site 2), and a homogeneous even-aged dense stand (site 3) were surveyed and photographed, and a 3D point cloud was extracted from the images. Then, the performance of algorithms was tested using a series of different canopy height models (CHM) with spatial resolutions of 25, 50, 75, 100, and 120 cm. To generate these models, digital surface model (DSM) was subtracted from digital terrain model (DTM). Results of individual tree delineation were validated using data collected in field observation of the aforementioned sites.
Results & Discussion
Results indicated that both RG and IWS algorithms yielded their best performance in the dense homogeneous structure. Moreover, the number of segments resulting from CHMs with low resolution was often much more than the actual number of trees. This was due to the occurrence of several peaks within an individual tree crown especially in low resolution images. With an F-score of 0.88, homogeneous even-aged dense stand (site 3) showed the highest overall accuracy in RG algorithm with a pixel size of 75 cm. Generally, results indicated that RG is an appropriate approach for individual tree delineation due to its flexibility in delineation of varying crown sizes. Furthermore, this method does not assume a circular shape for tree crowns and thus, is capable of detecting and segmenting irregular crowns. Generally, delineation of trees in urban forests using CHMs obtained from UAV-captured aerial imagery was highly accurate in homogeneous sites, while such models lacked efficiency in heterogeneous sites.
