【摘要】 针对一些织物图像,该文提出了一种新的基于角点分层匹配的图像拼接方法。该算法首先采用harris角点检测,然后用归一化互相关法进行初步的匹配,再运用稳健的RANSAC算法进行精匹配,最后对相关性最高的4对点求单应性矩阵,从而实现对图像的拼接。整个算法自动完成,采用二次匹配,鲁棒性强。实验结果表明,该方法能准确找到图像间的重叠位置,且融合效果比较理想。更多还原

【Abstract】 For some fabric images in this paper,a new image mosaic algorithm based on delamination matching of corner-points.In this algorithm,harris corner detector is used to extract feature points.Then,matching points are obtained though rough match using normalized cross-correlation and refine match using robust RANSAC.In the end,the image mosaic have achieved through the homography matrix calculated by 4 pairs of highest relativity matching points.The whole algorithm is completed automatically though twice matching with the high stability.Experimental results show that this algorithm can accurately find the location of overlap between images and the effect of amalgamation is comparative satisfied.更多还原