Research On Detection Method Of Pavement Texture Edge Distribution Uniformity Based On Two-dimensional Image Entropy

As a part of non-destructive testing,the detection of texture distribution uniformity plays an important role.However,current research has many shortcomings,as the detection results are severely affected by external factors such as lighting and shadows.Moreover,uniformity,as an indicator of uncertainty,lacks unified quantitative standards.In this study,asphalt concrete pavement images were used as experimental objects,and a complete image processing framework was constructed based on the twodimensional entropy of images to detect the uniformity of pavement texture.In addition,a new index called "uniform complexity(U)" was proposed to quantitatively evaluate the uniformity of texture distribution.First,the edge distribution map of pavement texture was extracted using image processing techniques.In the preprocessing of pavement images,the Otsu algorithm was employed for image binarization to separate the background and texture,thereby addressing the issue of uneven brightness and reducing the instability and errors in detection results.Different filtering algorithms were used to filter the binarized image in the image filtering stage,and a suitable filtering algorithm was selected through experimental comparison.Then,the edge segmentation algorithm based on the Canny operator was used to extract the edge distribution map of the image texture.Subsequently,the relationship between two-dimensional entropy and the uniformity of texture edge distribution was validated through simulation experiments.However,it was found that the detection results of two-dimensional entropy were not stable enough in the presence of image noise interference.Therefore,an image segmentation method was used to divide the global image into several sub-images,and a new metric called uniform complexity was proposed to quantify the uniformity of texture edge distribution.This metric is used to quantify the uniformity of texture edge distribution.Finally,to verify the applicability of the uniform complexity in practical detection,this paper applied the uniform complexity to detect the degree of pavement segregation and used the pavement instrument PQI380 as a comparative reference.Uniform complexity verification experiments and pavement noise experiments were designed.The results showed that uniform complexity can be used to detect the degree of asphalt pavement segregation,and the uniform complexity detection method is more stable than other methods for noisy images.