基于角点匹配的图像拼接技术  

Research of the Image Mosaic Method Based on Corner Match

作　　者：李冬梅[1] 王延杰[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所图像室

出　　处：《电子器件》2008年第3期919-922,共4页Chinese Journal of Electron Devices

摘　　要：采用的Harris角点匹配的算法是基于图像灰度的自相关函数,直接从灰度图像中检测兴趣点的算法。该算法简单,而且配准精度较高,具有较好的抗干扰的能力。此算法使图像拼接过程中对图像重叠律的要求降低到了15%,同时使图像的配准拼接速度提高了60%。达到了兼顾算法的精度和效率两个方面的目的。The Harris corner match algorithm in this paper is based on the correlation function of the gray image. It is an algorithm which detects interesting points directly from the gray image. The advantage of the algorithm is simpleness, precision location, and robustness to noise. This algorithm reduces 15% of overlapping areas of image for image mosaic. And it improves 600% of the image mosaic computation. So the proposed algorithm is both efficient and precise.

关 键 词：图像拼接 图像匹配 HARRIS算法 角点匹配 相关函数 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]