Texture Segmentation Applied to P. Oceanica Beds’ Upper Boundary Tracking by ROVs

The paper deals with an image processing technique for texture segmentation. By way of texture segmentation, a binary image is constructed and used to fit a line of the dominant direction of propagation of the texture in the image plane. The texture is captured by an oblique angle that positions the image plane in a near vertical orientation to the sea bottom plane. The texture being recognized in the experiment is that of Posidonion Oceanicœ, the benthic community (meadow or bed) of the Neptune grass Mediterranean endemic. It is the motivation of this paper for the dominant direction of propagation of the texture segment in the image plane to be used as a measure of the true direction of propagation of the upper border of Posidonion Oceanicœ in a feedback control loop that will enable the vehicle to autonomize the task of upper border tracking. The algorithm for extraction features four distinct phases: multi-resolution analysis using wavelets, vector quantization, post-processing of the obtained binary image and the extraction of the line parameters. The classification and line-fitting procedure are computationally optimized and made more robust by using weights in the Least Squares fitting procedure, and using nonlinear binary-image domain processing.