基于图像自动匀光的路面裂缝图像分析方法

A Pavement Crack Image Analysis Approach Based on Automatic Image Dodging

基于视频图像检测裂缝是当前路面病害检测的主要手段。为解决路面裂缝检测系统在不同光照条件下裂缝识别可靠性问题,研究了一种基于图像自动匀光的路面裂缝图像分析方法。首先对基于面阵CCD相机图像裂缝检测存在的问题进行分析,提出采用图像自动匀光技术解决不同光照条件下图像一致性输入问题;其次,设计了一种基于自动电子印相机原理的路面图像快速匀光算法,提出了一种实用的路面裂缝图像处理策略并设计了路面裂缝图像处理流程;最后,对一组由面阵CCD相机获取的路面图像按照该方法进行路面裂缝检测试验,验证了基于图像自动匀光的路面裂缝图像分析技术的合理性和实用性。

Image based crack detection technique is the current main method for pavement crack disease detection.In order to improve the reliability of the pavement crack recognition from image based pavement crack detection system under different illumination conditions,a pavement crack image analysis approach based on the image dodging was proposed.First,the problems of the current image based pavement crack detection system with planar array CCD camera were discussed.An automatic image dodging technique was presented to obtain the consistency input for image processing under different illumination conditions.Second,a fast image dodging algorithm based on automatic electric printer principle was designed and a practical intelligent pavement crack image processing strategy was proposed together with the pavement crack image processing flowchart.Finally,pavement crack image processing experiments for a set of images obtained by planar array CCD camera were implemented and analyzed with the presented approach to verify the reasonability and practicability of automatic image dodging based pavement crack recognition.