针对无人机影像匹配算法因影像信息量大、视场大和地物纹理复杂而导致匹配精度及效率低的问题,以加速稳健特征(Speeded Up Robust Features,SURF)算法为基础,提出了一种基于重叠区域改进SURF算法的无人机影像快速匹配算法.算法主要分为3步:①利用无人机影像重叠率计算影像特征点检测提取区域,并在区域内进行SURF特征点检测提取,避免无效特征点影响匹配精度与效率;②采用同心圆环8区域模板替换原SURF算法矩形模板,并引入影像色彩信息构建34维的特征描述符,进行特征匹配;③采用M估计抽样一致性(M-estimator Sample Consensus,MSAC)算法剔除误匹配点,实现影像的精匹配.实验分析表明,所提算法能够有效提升无人机影像特征匹配正确率及算法运行时间效率.
A Fast Matching Algorithm for Unmanned Aerial Vehicle Images Based on Overlapping Region Improvement SURF
In order to solve the problem of low matching accuracy and efficiency caused by the large amount of image information,large field of view,and complex terrain texture in UAV image matching algorithms,a fast matching algorithm for UAV images based on overlapping region improvement Speeded Up Robust Features(SURF)is proposed.The algorithm is mainly divided into three steps:①Using the overlapping rate of UAV images to calculate the image feature point detection and extraction area,and performing SURF feature point detection and extraction within the region to avoid invalid feature points from affecting matching accuracy and efficiency.②Replacing the original SURF algorithm rectangular template with a concentric ring 8-region template,and introducing image color information to construct a 34 dimensional feature descriptor for the feature matching.③ Using M-estimator Sample Consensus(MSAC)algorithm to eliminate mismatched points and achieving precise image matching.Experimental analysis shows that the proposed algorithm can effectively improve the accuracy of UAV image feature matching and the efficiency of algorithm runtime.
国家科技期刊平台
NSTL
万方数据
