Interactive texture segmentation via IT-SNAPS
Arunkumar Gururajan, Hamed Sari‐Sarraf, Eric Francois HEQUET · 2010
This paper presents a new framework for an interactive image delineation technique, which we term as the Interactive Texture-Snapping System (IT-SNAPS). The uniqueness of IT-SNAPS stems from the fact that it can effectively assist the user in accurately segmenting images with complex texture, without placing undue burden on the operator. This is made possible by unobtrusive extraction of information from the user during the delineation process, which in turn, is utilized on-the-fly to adapt IT-SNAPS to the boundary being segmented. The current work details (i) the mathematical construction of IT-SNAPS, (ii) demonstration of IT-SNAPS on a heterogenous texture image, and (iii) a performance comparison with the popular existing technique of intelligent scissors.