A morphological approach to pipe image interpretation based on segmentation by support vector machine

This paper presents a new approach to automatic pipe inspection using pixel-based segmentation of colour images by support vector machine (SVM) coupled with morphological analysis of the principal connected component of the segmented image. The pixel-based segmentation method has been tested using RGB, HSB, Gabor and local window feature sets and is seen to work best with the HSB feature set. The morphological analysis allows the principal connected component of the segmented image to be decomposed into the pipe flow line region, the pipe joints and adjoining defects. Generalisations of the morphological operations of erosion and dilation are defined and some simple properties of them are derived. A fuzzy approach to pipe connection detection is also described.