多边形形状分析的无人机影像重叠度计算方法

Method of UAV image overlap based on polygon shape analysis

针对无人机飞行姿态不稳定、地形起伏较大时,传统重叠度的统计方法计算不可靠、精度较差低、需要人工量测等问题,该文提出了一种多边形形状分析的重叠度统计方法。该方法首先利用基于区域特征与点特征相结合的高精度最小二乘匹配方法提取稳定可靠的影像同名点;然后使用改进的随机抽样一致性算法计算影像间的单应矩阵,并获取两影像的重叠多边形;最后采用多边形形状分析法获取最小重叠部分长度并计算得到重叠度。实验表明,该方法能够快速、准确地实现无人机影像重叠度的统计,与人工量测结果相比,误差均在3%以内,在实际生产中大幅提高生产效率和质检精度。

Due to the influence of terrain undulation and flight attitude,the traditional statistical method of overlap degree calculation is unreliable,the accuracy is low,and manual measurement is required,this paper proposes a statistical method of overlap degree based on polygon shape analysis.This method:first,use the high-precision least squares matching method based on the combination of regional features and point features to extract stable and reliable image match points;then,use an improved random sampling consensus algorithm to calculate the homography matrix between images,then get the overlapping polygons of the two images;finally,the polygon shape analysis method is used to obtain the length of the minimum overlapping part and calculate the overlap degree.Experiments show that this method can quickly and accurately realize the statistics of the degree of overlap of UAV images.Compared with the manual measurement results,the error is within 3%,which greatly improves the production efficiency and quality inspection accuracy in practical production.