基于模糊数学的公路路面裂痕检测技术-《集美大学学报（自然科学版）》

文章信息/Info

Author(s):: ?LIU Yi-ling; XIE Shu-tong; School of Computer Engineering,Jimei University,Xiamen 361021,China

Keywords:: pavement cracks; automatic detection; Fuzzy Mathematics; image segmentation

摘要:: 根据公路路面实际裂痕灰度值比背景灰度值低，而路面非背景和非裂痕像素灰度值比背景灰度值高或低很多，设计了一种基于模糊数学的公路路面裂痕检测的新技术.首先使用改进的基于乘性模型的灰度校正算法对原始图像进行预处理，接着根据OTSU最大类间方差法计算得到原图像前景与背景的最佳阈值作为渡越点，构造一个新的模糊隶属度函数对预处理后的图像进行模糊处理，然后用Sobel算子对模糊后的图像边缘检测判断得到含有裂痕信息的区域，最后用图像分割算法区域分割出含有裂痕信息的图像子块.实验证明，新提出的方法在处理速度和检测准确性方面基本能够满足实际需要，能实时、自动检测出模糊图像的裂痕.

Abstract:: With development of national economic and daily life of people,pavement cracking detection technology also needs to be automated,efficient and accurate.Based on the fact that the cracking pixel in pavement image are darker than background pixels,and non-background and non-cracking pixels are darker or brighter than background pixels,this paper designs a new method of pavement cracking detection.Firstly,the original image is preprocessed based on gray level correction algorithm of multiplicative models which we make some modification.Secondly,a transit point is got based on OSTU(maximum-between-clusters-variance method).Thirdly,a new membership function is constructed to blur the preprocessed image.Fourthly,the image containing cracks is detected by Sobel algorithm acting on the fuzzy image.Finally,the regions containing cracks are segmented out by image segmentation algorithm.The method proposed in this paper can basically meet the actual needs in processing speed and detection accuracy,and detect the cracks in fuzzy image in real-time and automatically.