摘要
本文提出了一种基于种子填充的图像二值化的算法,并将此算法应用于高速公路路面破损智能识别系统中,用来去除采集图像中的检测车影。实验表明此算法能改善目标与背景的分割效果,提高高速公路路面破损识别的正确率。
This paper describes a new seed filling based image thresholding algorithm, and this algorithm has been ap- plied to the system for recognition of freeway's surface diseases in order to eliminate the shadow of the detective bus from the gathered images. Experimental results show that this algorithm can improve the effect of segmentation of object and background and it improves the exactness of freeway's surface disease recognition.
出处
《计算机科学》
CSCD
北大核心
2005年第4期99-101,143,共4页
Computer Science
基金
江苏省自然科学基金(DK200208)
