摘要
针对无人机航拍图像拼接算法存在拼接速度慢、拼接结果不精准等问题,提出了一种新的无人机航拍图像快速拼接算法。采用A-KAZE算法构建非线性尺度空间并寻找局部极值点,利用改进的二进制特征描述子FREAK进行采样,采用主成分分析法对特征描述子进行精简降维。对最近邻比率策略设定搜索阈值,提高特征点配准效率,再利用随机样本一致性算法对特征点进一步提纯,使用加权平均融合算法对待拼接图像进行融合。实验结果表明,所提算法的特征点提取速度和特征点匹配速度较快,同时提高了特征点匹配正确率及拼接精度,拼接后的图像没有明显的拼接缝,整体颜色亮度保持一致,能够有效提高拼接图像速度和质量。
Aiming at the problems of slow stitching speed and inaccurate stitching result of UAV aerial photograph image stitching algorithm,a new fast stitching algorithm for UAV aerial photograph image is presented.A-KAZE algorithm is used to construct the non-linear scale space and find the local extreme points.The improved binary feature descriptor FREAK is used to sample,and the feature descriptor is simplified and reduced by principal component analysis.The search threshold is set for the nearest neighbor ratio strategy to improve the registration efficiency of feature points.Then,the feature points are further refined using the random sample consistency algorithm,and the stitched images are fused using the weighted average fusion algorithm.The experimental results show that the feature point extraction speed and feature point matching speed are faster,and the accuracy of feature point matching is improved.The stitched images do not have obvious stitching seams,and the overall color brightness is consistent,which can effectively improve the speed and quality of stitched images。
作者
李玉峰
任静波
LI Yu-feng;REN Jing-bo(School of Electronics and Information Engineering,Shenyang Aerospace University,Shenyang Liaoning 110136,China)
出处
《计算机仿真》
北大核心
2022年第5期201-207,共7页
Computer Simulation
基金
国家高分专项(70-Y40-G09-9001-18/20)
辽宁省自然基金项目(20180550334)
教育部科学技术研究重点项目(2017A0 2002)
辽宁省教育厅项目(L201701,L201735)。
关键词
无人机
航拍图像拼接
主成分分析法
Unmanned Aerial Vehicle
Aerial image stitching
Random sample consistency algorithm
