Geometric Calibration And Image Performance Evaluation Of Light And Small UAV Camera

With the rapid development of drone technology,drone aerial photogrammetry technology has become a powerful supplement to traditional aerial photogrammetry methods due to its low cost,high efficiency,and flexibility.In order to improve the accuracy of unmanned aerial vehicle aerial photogrammetry,it is necessary to perform geometric calibration on the non measuring cameras mounted on the drone to obtain internal parameters and distortion coefficients,and use them as a basis for aerial survey data processing in order to obtain better aerial survey results.The mainstream camera calibration method internationally is achieved through an outdoor calibration field,which meets the calibration requirements while also evaluating the performance of the image by establishing performance targets.However,most of the calibration fields at home and abroad are for satellite sensors,and few are for UAV aerial remote sensing integrated calibration fields.The current geometric calibration and performance evaluation methods also have some limitations,such as uneven illumination in color correction,poor noise and ambiguity resistance of traditional MTF measurement methods,and edge detection resolution to be improved,which need to be further improved.This article focuses on the above issues and combines domestic and foreign experience in the construction of calibration fields to establish calibration fields suitable for small and light unmanned aerial vehicles.Further research on camera geometric calibration and image performance evaluation is carried out,with a focus on analyzing and improving image performance evaluation methods.Finally,camera geometric calibration and image performance evaluation experiments are conducted based on the calibration field,verifying the reliability of the calibration field and the superiority of the algorithm proposed in this article.The main work and innovative points of the paper are as follows:(1)Aiming at the characteristics of small and light unmanned aerial vehicles(UAVs),considering both geometric calibration and image performance evaluation requirements,a calibration field suitable for small and light UAVs was established,and the reliability of the calibration field was verified and evaluated.(2)In terms of geometric calibration,based on the derivation of a computer vision model for beam adjustment,combined with the established calibration field and the Camera Calibration software developed by Jilin University,geometric calibration and image distortion correction are performed on the UAV camera.And study the impact of adding oblique images to the calibration process on the calibration results.(3)In terms of performance evaluation,resolution detection,color correction,and MTF measurement were mainly studied.Regarding resolution detection,image resolution detection is implemented based on three-line targets and radiation targets in the calibration field;Regarding color correction,the method of "first dodging,then correcting" is adopted.Through experimental comparison with common methods(no dodging,Mask dodging,Wallis dodging),the superiority of the proposed method of first using Retinex algorithm for image dodging,and then combining reference images for color correction is tested;Regarding MTF measurement,a new slanted-edge method based on improved Zernike moments for MTF measurement is proposed.Through simulation experiments and calibration field experiments,it is verified that the proposed algorithm is superior to traditional MTF measurement methods such as ISO 12233 and Hough transform.Using the edge target in the calibration field,the MTF of the camera was measured.(4)Development of image performance evaluation software for small and light unmanned aerial vehicles.Based on the established calibration field and related algorithms for image performance evaluation,a image performance evaluation software for small and light unmanned aerial vehicles(UAVs),Image Performance Evaluation1.0,was designed and developed based on MATLAB 2020 B,which can achieve performance evaluation such as image parameter calculation,resolution detection,and MTF measurement.In this paper,the establishment of a calibration field for small and light unmanned aerial vehicles and the analysis and improvement of geometric calibration and image performance evaluation methods are of great significance for improving the quality of photogrammetry and promoting the wide application of unmanned aerial vehicles in the field of surveying and mapping.