基于无人机测绘的复杂地形图像特征提取方法  

Complex terrain image feature extraction method based on drone mapping

作　　者：王亚妮 周银朋 WANG Yani;ZHOU Yinpeng(School of Biological and Envirronment Engineering,Xi’an University,Xi’an 721001,China;Surveying and Mapping Institute of Geology and Mineral Bureau of Guizhou Province,Guiyang 550018,China)

机构地区：[1]西安文理学院生物与环境工程学院,陕西西安721001 [2]贵州省地矿局测绘院,贵州贵阳550018

出　　处：《电子设计工程》2022年第19期149-152,158,共5页Electronic Design Engineering

基　　金：西安市科技计划项目(2020KJWL23)。

摘　　要：目前研究的地形图像特征提取方法存在灰度值与实际值相差较大,导致复杂地形图像特征提取精度较低等问题。为解决上述问题,提出基于无人机测绘的复杂地形图像特征提取方法。计算地形图像特征的多阈值,基于多阈值计算结果,提取地形图像特征点,通过SIFT按照递减顺序结合复杂地形图像特征点,完成匹配;采用隔阶比较的方法,对匹配后地形图像特征点进行处理,以此提升地形图像特征点提取的鲁棒性。实验结果表明,基于无人机测绘的复杂地形图像特征提取方法优于传统地形图像特征提取方法,其灰度值和实际值相差较小,吻合度较高,且提取精度较高。The current research on terrain image feature extraction methods has the problem of large difference between gray value and actual value,resulting in low accuracy of complex terrain image feature extraction. In order to solve the above problems,a feature extraction method of complex terrain image based on drone mapping is proposed. Calculate the multi threshold of terrain image features,extract the terrain image feature points based on the multi threshold calculation results,extract the combined complex terrain image feature points in descending order through SIFT extraction,and complete the matching;By using the method of step-by-step comparison,the matching terrain image feature points are processed,so as to improve the robustness of terrain image feature point extraction. The experimental results show that the complex terrain image feature extraction method based on drone mapping is better than the traditional terrain image feature extraction method. The difference between the gray value and the actual value is small,the coincidence degree is high,and the extraction accuracy is high.

关 键 词：无人机测绘 复杂地形 图像特征 特征提取 提取方法 图像提取 

分 类 号：P237[天文地球—摄影测量与遥感]