基于ORB特征点的道路图像拼接方法  

Road Image Mosaic Method Based on ORB Feature Points

作　　者：李明臻 姜梦炜 陈仕旗 LI Mingzhen;JIANG Mengwei;CHEN Shiqi(College of Engineering Machinery,Chang’an University,Xi’an 710064)

机构地区：[1]长安大学工程机械学院,西安710064

出　　处：《现代制造技术与装备》2022年第11期56-58,共3页Modern Manufacturing Technology and Equipment

摘　　要：为了快速获取更大范围且清晰度高道路图片,提出一种针对无人机(Unmanned Aerial Vehicle,UAV)近场采集的道路图像拼接方法。首先,在ORB(Oriented FAST and Rotated BRIEF)特征点提取的基础上,采用最邻近匹配算法进行特征点间的匹配。其次,通过汉明距离和随机采样一致(Random Sample Consensus,RanSaC)算法对匹配结果进行筛选,以获取准确的单应性矩阵。最后,采用最佳缝合线融合算法,使得图像过渡均匀。实验证明,所提方法可以有效处理无人机航拍路面图像,能够高效、准确地实现路面图像拼接。In order to quickly obtain a wider range of road images with sufficient clarity, a road image mosaic method for near field acquisition of Unmanned Aerial Vehicle(UAV) is proposed. Firstly, based on the extraction of Oriented FAST and Rotated BRIEF(ORB) feature points, the nearest neighbor matching algorithm is used to match the feature points. Then, through hamming distance and Random Sample Consensus(RanSaC) algorithm, the matching results are filtered to obtain an accurate homography matrix. Finally, the best suture fusion algorithm is used to make the image transition uniform. Experiments show that this method can effectively process UAV aerial road images, and can efficiently and accurately achieve road image mosaic.

关 键 词：ORB特征点 特征点匹配 随机采样一致(RanSaC) 图像拼接 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U418[自动化与计算机技术—计算机科学与技术]