Research on sonar image segmentation by spatial filling curve and grayscale distribution estimation

Aiming at the problem of target segmentation in sonar images, a method based on space filling curve and grayscale distribution estimation is proposed and named as SFC-GDE Method. In this method, the space filling curve is firstly used to transform sonar image from two-dimension matrix to one-dimension vector, and then, the entire sonar image processing process, including filtering, grayscale distribution estimation, threshold calculation and image segmentation, are carried out in one dimensional space. Finally, the segmented one-dimensional vector is transformed into a two-dimensional matrix by inverse transformation to achieve target segmentation. Actual sonar image has been processed to verify the effectiveness of the proposed method.