收稿日期:2006-11-29
基金项目:特殊地区公路工程教育部重点实验室资助项目(KLP200501)
作者简介:闫茂德(1974-),男,陕西澄城人,副教授,主要研究方向为数字图像处理和非线性系统控制等。
2008 年 工 程 图 学 学 报 2008
第 2 期 JOURNAL OF ENGINEERING GRAPHICS No.2
一种基于形态学的路面裂缝图像检测与分析方法
闫茂德,
伯绍波,
贺昱曜
(长安大学信息工程学院,陕西 西安 710064)
摘 要:基于数学形态学的原理,提出了一种路面裂缝图像检测与分析的新方法。
该方法通过 4 种结构元素的中值滤波算法增强路面灰度图像,并利用形态学梯度算子、闭合
算子进行边缘检测和裂缝空隙弥合,能够有效地检测出裂缝边缘,并保持边缘的平滑性。在
此基础上,利用改进的腐蚀算子细化裂缝并获取其骨架,通过计数的方法得到裂缝的长度和
宽度。实验结果表明,该方法有抗噪性能好、边缘提取能力强、路面裂缝特性分析精确和计
算量较小等优点。
关 键 词:计算机应用; 裂缝检测; 形态学; 图像处理
中图分类号:TP 391.41
文献标识码:A 文 章 编 号:1003-0158(2008)02-0142-06
A Method of Image Detection and Analysis for
Pavement Crack Based on Morphology
YAN Mao-de, BO Shao-bo, HE Yu-yao
( School of Information Engineering, Chang’an University, Xi’an Shanxi 710064, China )
Abstract: In order to detect the edge efficiently and get the characteristics of pavement
cracks exactly, a novel image processing method based on the theory of the morphology is
presented for the pavement crack detection and analysis. The method employs median-filter
algorithm with the four structural elements to enhance the gray-scale pavement images, and
combines morphology-gradient operator with close operator to extract the crack edges and fill the
gaps of the cracks. After the above processing, the presented method utilizes the improved erosion
operator to thin the cracks and obtain the skeletons, and then gets the length and width of cracks
by counting. Experimental results indicate that the proposed method has good performance of
noise reduction, strong ability of extracting edges, exact analysis of pavement crack and fewer
calculations.
Key words: computer application; crack detection; morphology; image processing