UAV image matching based on improving the ORB algorithm

doi:10.13474/j.cnki.11-2246.2024.0121

Abstract

Abstract: This paper addresses the problem of unstable feature point extraction and pixel-level feature point localization accuracy faced by the ORB algorithm when encountering changes in illumination. To improve the ORB algorithm, a self-adaptive threshold method based on foreground-background contrast is proposed in conjunction with existing sub-pixel localization techniques. In order to avoid the error caused by the need for manual threshold setting in RANSAC algorithm, this paper introduces the MAGSAC++algorithm into the feature matching process for false match elimination. Experimental results show that the improved algorithm can obtain a larger number of matches, has better robustness to changes in illumination, and improves matching accuracy by at least 7%.

Key words: ORB algorithm, feature extraction, adaptive threshold, image matching, mismatch removal

CLC Number:

LI Bing, LAI Zulong, SUN Jie, DING Kaihua. UAV image matching based on improving the ORB algorithm[J]. Bulletin of Surveying and Mapping, 2024, 0(1): 126-130,149.

References

[1] 陈晓勇,何海清,周俊超,等. 低空摄影测量立体影像匹配的现状与展望[J]. 测绘学报,2019,48(12): 1595-1603.
[2] SHARMA S K,JAIN K. Image stitching using AKAZE features[J]. Journal of the Indian Society of Remote Sensing,2020,48(10): 1-13.
[3] ZHANG Jin,YU Huan,DENG Huaxia,et al. A robust and rapid camera calibration method by one captured image[J]. IEEE Transactions on Instrumentation and Measurement,2019,68(10): 4112-4121.
[4] 周祖鹏,张晓东,甘良棋. 无人机航拍图像的三维重建方法研究[J]. 机械设计与制造,2021(5): 47-50.
[5] 娄佩卿,陈晓雨,王疏桐,等. 基于无人机影像的喀斯特农耕区地物识——以桂林市为例[J]. 国土资源遥感,2020,32(1): 216-223.
[6] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision,2004,60(2): 91-110.
[7] BAY H,TUYTELAARS T,GOOL L V. Surf: Speeded up robust features[C]//Proceddings of 2006 European conference on computer vision.[S.l.]:Springer,2006.
[8] ROSTEN E,DRUMMOND T. Machine learning for high-speed corner detection[M]//Computer Vision-ECCV 2006. Berlin,Heidelberg: Springer Berlin Heidelberg,2006: 430-443.
[9] RUBLEE E,RABAUD V,KONOLIGE K,et al.ORB:an efficient alternative to SIFT or SURF[C]//Proceedings of 2011 International Conference on Computer Vision.Barcelona:IEEE,2012:2564-2571.
[10] 李国竣,徐延海,段杰文,等. 利用局部自适应阈值方法提取ORB-SLAM特征点[J]. 测绘通报,2021(9): 32-36.
[11] 党宏社,李俊达,张选德. 基于加权网格运动统计的改进ORB算法[J]. 陕西科技大学学报,2022,40(1): 182-187.
[12] 宗泽华,张海军,张劲锋.基于自适应阈值的ORB特征提取与跟踪[J].电光与控制,2022,29(3):110-114.
[13] LI Shiji,WANG Qing,LI Jiayue.Improved ORB matching algorithm based on adaptive threshold[J].Journal of Physics:Conference Series,2021,1871(1):012151.
[14] DAI Yong,WU Jiaxin. An improved ORB feature extraction algorithm based on enhanced image and truncated adaptive threshold[J]. IEEE Access,2023,11: 32073-32081.
[15] BARÁTH D,NOSKOVA J,IVASHECHKIN M,et al. MAGSAC,a fast,reliable and accurate robust estimator[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle,WA:IEEE,2020: 1301-1309.
[16] FISCHLER M A,BOLLES R C. Random sample consensus[J]. Communications of the ACM,1981,24(6): 381-395.
[17] CALONDER M,LEPETIT V,STRECHA C,et al. BRIEF: binary robust independent elementary features[M]//Computer Vision-ECCV 2010. Berlin,Heidelberg: Springer Berlin Heidelberg,2010: 778-792.
[18] OTSU N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1): 62-66.
[19] 程希希,张艳玲,田军委. 基于模板匹配的快速角点检测方法[J]. 计算机工程,2020,46(7): 222-227.
[20] 谷学静,马冠征,周士兵,等.融合边缘化外点处理的改进图像匹配方法[J].激光杂志,2022,43(3):82-86.
[21] 闫利,叶志云.几何约束条件下的SIFT倾斜影像匹配[J].测绘通报,2016(1):37-40.