|
On detection and representation of multiscale low-level image structure Discusses the requirements of a computer vision system. A capability for bottom-up data reduction is necessary. The low level structure would serve as a lossless image abstraction and help initiate hierarchical, closed-loop image interpretation, for example, for recognition by enforcing a priori semantic constraints involving part-whole relationships. This note is not concerned with interpretation processes; it describes some desirable characteristics of strategies for the detection and representation of low-level perceptual structure or multiscale segmentation, which remains an open problem. Keywords: computer vision; pattern recognition; image representation; multiscale low-level image structure; bottom-up data reduction; data compression; lossless image abstraction; low level structure; hierarchical closed-loop image interpretation; image recognition; a priori semantic constraints; low-level perceptual structure; multiscale segmentation. Full Text: PDF |