基于扫描线分段的图像稠密匹配方法

Image Dense Match Method Based on Scanline Segmention

在分析相关视图扫描线特征的基础上,提出了一种基于扫描线分段图像稠密匹配方法。依据灰度特征,将校正图像扫描线划分为灰度相似的像素段,重构以灰度平均的像素段为基元的新扫描线,在计算像素段匹配的基础上进行图像稠密匹配,以降低视图稠密匹配过程中的歧义性,提高匹配效率,最后给出了实验结果。

This paper proposes an image dense match method using scanline segmentation based on the analysis of pixel intensity on scanline. The scanline on rectified images are divided into gray-consistent segments, each segment is filled with average gray pixel to form the constructed scanlines. The dense match are computed on the constructed scanlines. The experiments prove that the dense match efficiency is improved while fuzziness is reduced.