Automatic segmentation of bleeding regions in asphalt pavement using multiresolution analysis of the image

Pavement is the main part of the country's road network. Therefore, the evaluation of pavement conditions, especially the inspection of pavement distresses has an important role in the managing and repairing process of the pavement network. Bleeding is one of the important distresses in asphalt pavement because it has a direct influence on pavement skid resistance, vehicle maneuverability, and road safety. According to a survey of prior research, the automatic evaluation of bleeding receives less attention than other pavement distresses, and the main focus of study in this field has been on the automatic detection of bleeding. Automatic segmentation of distress regions is a critical component of pavement distress evaluation, allowing for a more precise assessment of distress. As a result, the purpose of this study is to develop an efficient method for the automatic segmentation of bleeding regions from healthy regions using image processing tools based on multiresolution image analysis. The proposed method has better performance than prior research, with averages of 82.44, 88.66, and 80.33 percent based on recall, Dice similarity coefficient, and intersection to union ratio as prevalent segmentation performance metrics. This model allows for the automated evaluation of bleeding and can enhance the performance of pavement management systems.