An improved multiresolution technique for pavement texture image evaluating

The texture assessment of pavement surface plays an important role in nearly all Pavement Management System (PMS). The Surface drainage, which related to the surface texture condition and plays a vital role in pavement safety and accident rate reduction, is a generally neglected part in these systems. The present study conducted to present an image-based method for assessment of the surface drainage capability of pavements. The presented method consists of three steps: image preprocessing, feature extraction and evaluation of pavement surface drainage. In the first step, a Modified Shearlet Transform (MST) with specific filter and parameters is presented for image denoising and then a set of morphological features are extracted to present proper evaluation parameters for surface drainage assessment. Finally, a method based on feature selection for surface drainage evaluation proposed and a classification for the pavements based on surface drainage quality by the C5.0 algorithm conducted considering the extracted parameters. Experimental results demonstrate that the MST is fast and efficient for evaluation of pavement surface drainage.

Automated summary

The assessment of surface drainage is an important aspect in evaluating pavement condition, however it is often neglected in existing pavement management systems. This study proposes an image-based method for evaluating surface drainage capability of pavements, which consists of three steps: image preprocessing, feature extraction, and evaluation. Experimental results show that the proposed method is fast and efficient for assessing pavement surface drainage.