摘要
提出了一种图像插补的方法。首先将两幅图像进行矫正,使得它们的对极线处在同一扫描线上,从而将二维图像的匹配简化成一维图像的匹配,使得下一步的匹配快速、准确;然后检测第一幅图像扫描线上的边缘点,利用这些边缘点分割两幅矫正图像对应的扫描线,进行密集匹配;最后确定插补图像的像素位置和灰度。实验证明该算法简单,容易实现。
A method named image interpolation is given in this paper. First the two images are rectified to make their epipolar lines parallel to the scan- lines. In this way,the matching of 2D images is simplified to 1D,so image matching is made quickly and nicely. Next,matching points are searched in the new images. With these corresponding edge points, the scan- lines of the two rectified images are segmented and dense matching is run. Position and gray value of pixes in the image are determined in the end. This algorithm is simple and easy to realize which has been proved by the experiments.
出处
《计算机技术与发展》
2006年第6期130-131,135,共3页
Computer Technology and Development
关键词
图像矫正
图像插补
密集匹配
扫描线
image rectify
image interpolation
dense matching
scan- lines
