基于曲张特征的快速Hough逆变换算法

提出一种基于曲线特征(CF)的Hough逆变换(IHT)算法.基于静止图像中待检测曲线的特征,选取图像中若干特征点,根据这些特征点,结合待检测曲线的特征信息寻找生长点,以这些生长点为下一次检测过程的新特征点,循环生长,直至达到检测要求为止.实验表明,CF-IHT算法只选取了图像中待检测曲线上的点进行重建,克服了标准IHT(SIHT)逐像素检测的盲目性,减小了算法的计算复杂度,节省了存储空间,提高了算法的计算效率.

A Fast Inverse Hough Transform Based on Curve Feature

A new inverse Hough transform based on curve feature(CF-IHT) algorithm was presented.Based on the feature of the curve to be detected,the new algorithm selects some feature points in the edge image.According to the coordinates of the feature points and the CF,it finds some grew points.The grew points will be the new feature points in the next process of finding the other new grew points.This process will be end until the detecting result meets the established criteria.Experimental results show that the CF-IHT is less blind and reduces the computation complexity and storage burden of standard IHT(SIHT).