摘要
研究煤矿井下图像拼接的算法,探讨井下环境光照不均、强噪声的图像拼接问题。用当前流行的四种特征点检测算法,如Moravec算子、Forstner算子、Harris角点检测算法和CSS角点检测算法提取井下图像的特征点,分析得出CSS角点检测算法更适合井下工作环境的结论。用CSS角点检测算法提取待拼接图片特征点,通过归一化互相关系数(NCC)提取特征点匹配对,并通过RANSAC删除伪匹配对,最终实现无缝拼接。
Studies the mosaic algorithm for weak,uneven ilumination and strong noise images from coal-mine underground.Extracts the characteristic points of underground image by some recently popular detection algorithms.Analysis comes to the conclusion that the CSS corner detection algorithm is the most suitable one to miine underground environment.Extracts the characteristic points by this algorithm and treats by NCC（Normalized Cross Correlation）.Funthermore,deletes false matching points by RANSAC and obtain the seamless mosaic image
出处
《煤矿机电》
2011年第3期41-43,46,共4页
Colliery Mechanical & Electrical Technology
关键词
图像拼接
特征点检测算法
CSS角点检测算法
匹配
image mosaic
detection algorithm of characteristic point
CSS corner detection algorithm
match
