Target detection The Simulation Analysis of the Effect of Imaging Sidelap/Overlap on the Quality of 3D Reconstruction in UAV Photogrammetry and Determination of their Optimal Values
|
Ali Erfanzadeh * , Mohammad Saadatseresht |
University of Tehran |
|
Abstract: (2540 Views) |
The selection of effective and optimal sidelap/overlap of images according to the quality, speed, cost and purpose of each UAV photogrammetric project is considered as an important parameter. Since two decades ago, several researches have been carried out based on a trial and error strategy for this purpose, in which different drones and cameras have been utilized for imaging with different sidelaps/overlaps and analyzing the results obtained from them. Due to their high cost and time-consuming these operational methods, have not yet been able to provide a perfect behavioral assessment of the impact of imaging sidelap/overlap parameters on the quality of three-dimensional reconstruction. These studies merely provide an overview of the sidelap/overlap of images between 60 and 80% empirically. In this paper, a simulation method is used to comprehensively evaluate the effect of the imaging sidelap/overlap parameters on the quality of three-dimensional reconstruction in UAV photogrammetry. For this purpose, due to the complexity of the subject and to simplify it, the problem of three-dimensional reconstruction of a hypothetical object point has been analyzed by Monte Carlo analytical simulation using vertical aerial images with different sidelaps/overlaps. The results of the performed experiments in five modes of ideal, excellent, good, medium and bad respectively , show that changing the sidelap/overlap of images can improve the 3D reconstruction error for RMSExy up to 10 times and for RMSExyz up to 5 times. It can also be said that to achieve a higher quality of three-dimensional reconstruction, when the total sidelap/overlap of images is more than 120%, it is better to increase the coverage difference, and when it is less than 120%, it is better to reduce it. In addition, if you want to choose one of sidelap or overlap below 50%, leave the another one between 85% and 95%, or select both above 50% to keep the quality of the 3D reconstruction. In this research, for the first time, a new four step process on the basis of the Pareto front algorithm was introduced in order to identify and select the optimal imaging sidelap/overlap. |
|
Keywords: sidelap/overlap of images, 3D reconstruction, UAV photogrammetry, Simulation, Monte Carlo. |
|
Full-Text [PDF 2896 kb]
(1172 Downloads)
|
Type of Study: Research |
Subject:
Aerial Photogrammetry Received: 2021/03/18 | Accepted: 2021/07/19 | Published: 2021/07/22
|
|
|
|
|
Send email to the article author |
|