Solving discontinuity and distortion problems in boundary based image segmentation and representation

The boundary based image segmentation and representation system developed by Nobors and Ranganath has been found suitable for a wide range of image interpretation applications [10]. The segmentation system receives a thinned edge image as input and outputs a polyline for each open curve, and a polygon for each closed curve in the image. A novel representation scheme which uses both graph and tree data structures provides flexibility and permits hierarchical interpretation of the image. However, the approach does not address the effects of boundary discontinuity and distortion on the system's performance. This paper discusses these ill-effects and presents computationally efficient methods to bridge discontinuities and smooth distortions by directly post-processing the graph/tree hierarchical data structure that contains the boundary information produced by the segmentation system.

[1]  Henri Briand,et al.  Shape representation for image retrieval , 1999, MULTIMEDIA '99.

[2]  Zen Chen,et al.  Recursive conversion of chain code into quadtree with table lookup , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[3]  H. S. Ranganath,et al.  A boundary-based image segmentation and representation method for binary images , 2000 .

[4]  Dana H. Ballard,et al.  Generalizing the Hough transform to detect arbitrary shapes , 1981, Pattern Recognit..

[5]  Paul Wintz,et al.  Digital image processing (2nd ed.) , 1987 .

[6]  Josef Kittler,et al.  The Adaptive Hough Transform , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Chaur-Heh Hsieh,et al.  Coding of arbitrarily shaped objects using pace code , 2000, IWISPA 2000. Proceedings of the First International Workshop on Image and Signal Processing and Analysis. in conjunction with 22nd International Conference on Information Technology Interfaces. (IEEE.

[8]  Hanan Samet,et al.  A differential code for shape representation in image database applications , 1997, Proceedings of International Conference on Image Processing.

[9]  Subramaniam Ganesan,et al.  A robust Hough transform technique for description of multiple line segments in an image , 1998, Proceedings 1998 International Conference on Image Processing. ICIP98 (Cat. No.98CB36269).