Feature extraction of sewer pipe failures by wavelet transform and co-occurrence matrix

Traditionally, the sewer inspection usually discovers sewer failures on numerous CCTV images by human interpretation. However, it remains to be improved in both consideration of economic and efficient due to human's fatigue and subjectivity. To enhance the sewer inspection approaches, this paper attends to employ artificial intelligence into image process to extract the failure features of the sewer systems, which was also applied to the sewer system in the eastern Taichung City, Taiwan. The extracted features are valuable information in pattern recognition of failures on CCTV images. Wavelet transform and gray-level co-occurrence matrix, which have been widely applied in many texture analyses. were adopted in this research. Wavelet transform is capable of dividing an image into four sub-images including approximation sub-image, horizontal detail sub-image, vertical detail sub-image, and diagonal detail sub-image. In this paper, the co-occurrence matrixes of horizontal orientation, vertical orientation, and 45° and 135° orientations, respectively, were calculated for the horizontal, vertical, and diagonal detail sub-images. Subsequently, the features including angular second moment, entropy, contrast, homogeneity, dissimilarity, correlation, and cluster tendency, can be obtained from the co-occurrence matrixes. However, redundant features either could decrease the accuracy of texture description or could increase the difficulty of pattern recognition. Thus, the correlations of the features are estimated to find out the appropriate feature sets in which the coefficients of correlation of the features are less than 0.5. Finally, a discriminant analysis was used to evaluate their discriminabilities to the pipe defect patterns, and entropy, correlation, and cluster tendency were the best feature vector because of its better discriminant accuracy according error matrix analysis.