中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊
重庆交通大学学报（自然科学版）

重庆交通大学学报（自然科学版） ›› 2010, Vol. 29 ›› Issue (1): 69-72.

• • 上一篇    下一篇

基于小波的路面裂缝识别研究

孙波成1,2，邱延峻1，梁世庆3   

  1. 1.西南交通大学土木工程学院，四川 成都 610031;2.西南交通大学峨眉校区，四川 峨眉 624202; 3.襄樊学院，湖北 襄樊 441023
  • 收稿日期:2009-10-14 修回日期:2009-10-28 出版日期:2010-02-15 发布日期:2015-01-22
  • 作者简介:孙波成(1964—)，男，四川省孝感人，博士，主要从事道路工程和计算机应用研究。E-mail:sunboch@163.com。
  • 基金资助:
    交通运输部西部交通建设科技项目(2004 318 00054);四川省交通科技项目(SC200600382C)

Cracking Recognition of Pavement Surface Based on Wavelet Technology

SUN Bo-cheng1,2，QIU Yan-jun1，LIANG Shi-qing3   

  1. 1.School of Civil Engineering，Southwest Jiaotong University，Sichuan Chengdu 610031，China; 2.Emei Campus of Southwest Jiaotong University，Sichuan Emei 614202，China; 3.Xiangfan Institute，Hubei Xiangfan 441023，China
  • Received:2009-10-14 Revised:2009-10-28 Online:2010-02-15 Published:2015-01-22

PDF

摘要： 应用基于小波变换的图像分解和重构方案可以进行公路路面裂缝的检测与识别。首先对路面图像进行多分 辨二维离散小波分解，根据在每个尺度2j上小波系数模最大值和相位(梯度方向)这两个分量的大小确定边缘的 位置，进一步变换小波分解系数，达到去除噪音，加强裂缝边缘的效果。并通过实验分析和讨论了不同的小波基、 不同分辨率水平、重构中使用的分解子图像对路面图像裂缝分割的影响，最后用简单的阈值技术生成二值图像，把 裂缝从背景图像中分离出来。

关键词: 道路工程, 裂缝识别, 小波变换, 路面裂缝, 图像重构, 图像分割

Abstract: The image decomposition and reconstruction programs based on wavelet transform are applied to the detection and recognition of pavement cracking. Firstly，2-D discrete wavelet multi-resolution decomposition is carries out on the image of the pavement. The location of the edge is determined by the size of two components，that is，the phase (gradient direction) and local modulus maxima of the wavelet transform at different decomposition levels. The wavelet decomposition coefficients are further transformed to remove noises and to enhance the edges of cracks. Then，effects of different wavelet bases， different multi-resolution levels，the decomposed sub-images used in reconstruction on the segment of pavement image cracks are discussed and analyzed through experiments. Finally，simple threshold technology is used to generate binary image， which can discriminate cracks against background regions in the reconstructed images.

Key words: road engineering, cracking recognition, wavelet transform, pavement cracking, image reconstruction, image segmentation

中图分类号: