Application of Shape From Shading Technique for Side Scan Sonar Images

Digital signal processing technology has revolutionized a way of processing, visualisation and interpretation of data acquired by underwater systems. Through many years side scan sonars were one of the most widely used imaging systems in the underwater environment. Although they are relatively cheap and easy to deploy, more powerful sensors like multibeam echo sounders and sonars are widely used today and deliver 3^D bathymetry of sea bottom terrain. Side scan sonar outputs data usually in a form of grey level 2^D acoustic images but the analysis of such pictures performed by human eye allows creating semi-spatial impressions of seafloor relief and morphology. Hence the idea of post-processing the side scan sonar data in a manner similar to human eye to obtain 3^D visualisation.

In recently developing computer vision systems the shape from shading approach is well recognized technique. Applying it to side scan sonar data is challenging idea used by several authors. In the paper, some further extensions are presented. They rely on processing the backscattering information of each footprint (pixel in sonar image) along with its surroundings. Additionally, a current altitude is estimated from the size of shadow areas. Both techniques allow constructing 3^D representation of sea bottom relief or other investigated underwater objects.