Road Surface Crack Detection Based On Local Mean Standard Deviation Algorithm

With the continuous development and improvement of our country’s economic strength,the number of road,traffic volume and the pressure of the road are also increasing,which makes the road maintenance work intensity increase.Therefore,it is more and more concerned about the highway pavement disease detection in a timely and fast way.Because of the type of disease is a kind of initial disease,if the maintenance department is not timely,will be developed into other more serious type of road disease.Therefore,how to accurately and quickly detect the road surface cracks is an urgent problem to be solved.In this paper,after the first step analysis,the quantitative analysis of the identification process.According to the noise appearing in the image of the road surface,the preprocessing is carried out,and the mean value filter is used to process the image,the adaptive Wiener filter and Wallis transform are used to enhance the image.For the extraction of the internal characteristics of the pavement image binarization algorithm,using P-tile processing results of pavement image,and extract the characteristic parameters of the pavement image,and then use artificial neural network to classify pavement image.In order to solve the uneven distribution of gray level distribution in the surface of the cracks in the road surface,the Mask smoothing method is adopted.There are a lot of data in the road surface image,and the block processing technique is adopted,and then the local mean value standard deviation algorithm is used to identify the cracks.In order to evaluate the damage degree of the crack,the total area of the image block and the area of the minimum external rectangle need to be calculated.The experimental results show that this method can not only improve the detection efficiency of pavement crack diseases,but also can accurately mark the location of cracks.