Sonar Image Clustering Analysis by SOFM

Seabed image is clustered according to SOFM network. The feature vectors are average intensity, six statistics of texture and two dimensions of fractal. It takes the spatial correlation between different pixels and the terrain coarseness into consideration. Double blanket algorithm is used to calculate dimension. Because a uniform fractal may not be sufficient to describe a seafloor, two dimensions are calculated, respectively, by the upper blanket and the lower blanket. Dimensions are different in across-track and along-track, so the average of four directions is used to solve this problem.