Abstract
In this paper, a single-pass algorithm for extracting 8-direction chain-coded contours from a binary image, and determining contours inclusion relationship is described. The algorithm processes the image using a 2×2 pixels window sliding in a raster scanning fashion. By observing the sequence of patterns through this window, intermediate structures called fragments may be created, lengthened, merged, or closed to form a complete contour. Furthermore, inclusion relationships among fragments and/or contours are maintained throughout the processing, and are operated in such a way that once a contour is formed, its relationship with those contours being enclosed must be correct. Realization of the algorithm on a sequential machine is compared with other algorithms in the literature showing that for images with more contour points, the proposed algorithm is the fastest.
Original language | English |
---|---|
Title of host publication | IEEE Pac Rim Conf Commun Comput Signal Process |
Publisher | Publ by IEEE |
Pages | 256-259 |
Number of pages | 4 |
ISBN (Print) | 0780312198 |
Publication status | Published - 1 Jan 1993 |
Event | Proceedings of the IEEE 1993 Pacific Rim Conference on Communications, Computers and Signal Processing - Victoria, BC, Canada Duration: 19 May 1993 → 21 May 1993 |
Conference
Conference | Proceedings of the IEEE 1993 Pacific Rim Conference on Communications, Computers and Signal Processing |
---|---|
Country/Territory | Canada |
City | Victoria, BC |
Period | 19/05/93 → 21/05/93 |
ASJC Scopus subject areas
- Signal Processing
- Computer Networks and Communications