Motion estimation and tracking of multiple objects in sector scan sonar using optical flow

The new generation electronic sector scanning sonars is now allowing practical use of temporal information for signal processing tasks such as object classification and motion estimation. This paper presents an approach to the segmentation, 2D motion estimation and subsequent tracking of multiple objects in sector scan sonar image sequences. Initially, frequency domain filtering is used to distinguish static and moving objects in the sonar scan image sequence. Optical flow calculations are then performed on moving objects with significant size, to obtain magnitude and direction motion estimates. Matching of these motion estimates, and the object future positions they predict, is then used as a basis for identifying corresponding objects in adjacent scans. A tracking tree is constructed storing multiple possible correspondences and cumulative confidence values obtained from successive compatibility measures. Deferred decision making is then employed to enable best estimates of object tracks to be updated as subsequent scans produce new information.