Analyzing the Degree of Wavelet Directional Sensitivity in Smart Recognition of Asphalt Pavement Distress Texture

Evaluation of the road surface distresses is one of the most prominent phases of pavement management process. Over the past decade, a considerable number of researches have been carried out on developing automatic methods for distress detection most of which rely on computer vision and image processing techniques. One of the most important assets comprising computer vision systems is the image feature extraction algorithm. Textural features present more detailed information about the image regions characteristics compared to other features such as color and geometrical (shape) properties. In the past few years, multi-resolutional analysis approaches, such as wavelet transforms have provided an effective tool for fast and accurate image texture representation. In the present study, after acquisition of six different types of asphalt pavement distresses under controlled condition, in order to analyze their structures, three 2-D discrete wavelet transforms including Haar, Daubechies3 and Coiflet1 were utilized. In addition to aforementioned transforms, directional selective dual-tree complex wavelet transform was also applied on the distress images with the purpose of investigating the effectiveness of directional sensitivity increasement. After decomposition of the images by applying the abovementioned transforms, second-order statistics based on gray level co-occurrence matrix and higher-order descriptors based on gray level run-length matrix were employed, in order to characterize the wavelet frequency sub-bands texture.