实时图像处理中一种快速的直线检测算法

The replace algorithm of hough transform in real-time image processing

Hough变换(HT)是一种常用的直线检测方法.由于它具有很高的计算复杂度,在既无HT专用芯片也无阵列处理器的环境下,难以满足实时图像处理系统对实时性的要求.作者对此问题进行了研究,提出了一个新的快速算法——BL(Blob-basedLinedetection)算法,在普通计算机环境下代替HT进行直线检测.该算法突破了HT对实时系统造成的速度瓶颈,完满地解决了这个问题.

Hough Transform(HT), is a traditional method used to extract lines. Because of its high consumption of computations, it cannot meet the requirement of real\|time image processing if no special chip for HT or no array of processors in the processing system. We suggest Blob\|based Line\|dection(BL) algorithm, a new and high performance algorithm, to replace the HT in conventional computer environment. The new method breaks the bottleneck that HT has made, and has been used successfully in a real\|time image processing system.