Abstract: In this paper, the Beamlet transformbased algorithm is proposed to improve the reliability of the pavement crack detection under different illumination conditions.The algorithm divides the picture into blocks by the binary recursive method.The variety scales Beamlet is used to restore the crack, and thus the cracks are separated from the background picture.The interpolation function is used to dodge the background.In process of crack detection, this method filters the noise, and improves the accuracy of the extracted crack.Compared with the traditional edge detection algorithm, the proposed algorithm not only effectively restrains noise jamming but accurately detects the road crack, with fast computing speed.Uniform background lay the foundation for the qualitative and quantitative analysis of pavement crack image. 

Key words: intelligent transportation, image dodging, Beamlet transform, pavement crack detection, pixel gray, noise

摘要： 为解决路面病害检测过程中在不同光照条件下裂缝识别可靠性的问题，Beamlet变换提供了一种多尺度分析算法．该算法采用递归二进法将图像分块，利用在所有像素尺度下的Beamlet线段逼近裂缝，将裂缝从图像背景中分离，并采用插值函数均匀光照背景．在图像裂缝提取过程中，对图像中裂缝的线性提取可以有效地滤除噪声，提高裂缝提取的准确性．较传统的图像处理算法，该算法在准确地检测路面裂缝，均匀光照背景的同时有效地抑制了噪声干扰，检测效果良好，且运算速度快．均匀的路面背景为路面裂缝图像的定性、定量分析奠定基础．

关键词: 智能交通, 匀光, Beamlet变换, 路面裂缝检测, 像素灰度, 噪声