Sidescan sonar segmentation using active contours and level set methods

This paper is concerned with the application of active contour methods to unsupervised binary segmentation of high resolution sonar images. First texture features are extracted from a side scan image containing two distinct regions. A region based active contour model of Chan and Vese [3] is then applied to the vector valued images extracted from the original data. The set of features considered is the Haralick feature set based on the co-occurence matix. To improve computational efficiency the extraction of the Haralick feature set is implemented by using sum and difference histograms as proposed by Unser [16]. Our implementation includes an automatic feature selection step used to readjust the weights attached to each feature in the curve evolution equation that drives the segmentation. Results are shown on simulated and real data. The influence of the algorithm parameters and contour initialisation are analysed. ©2005 IEEE.