Computing Distance Maps Efficiently Using An Optical Bus
This paper discusses an algorithm for finding a distance map for an image efficiently using an optical bus. The computational model considered is the arrays with a reconfigurable pipelined bus system (LARPBS), which is introduced recently based on current electronic and optical technologies. It is shown that the problem for an n × n image can be implemented in O(log n log log n) time deterministically or in O(log n) time with high probability on an LARPBS with n 2 processors. We also show that the problem can be solved in O(log log n) time deterministically or in O(l) time with high probability on an LARPBS with n 3 processors. The algorithm compares favorably to the best known parallel algorithms for the same problem in the literature.
KeywordsBlack Pixel Processor Array Left Region Information Processing Letter Euclidean Distance Transformation
Unable to display preview. Download preview PDF.
- 4.D. Chiarulli, R. Melhem, and S. Levitan, “Using Coincident Optical Pulses for Parallel Memory Addressing,” IEEE Computer, vol. 20, no. 12, pp. 48–58, 1987.Google Scholar
- 7.Yi Pan and Keqin Li, “Linear array with a reconfigurable pipelined bus system: concepts and applications,” Special Issue on “Parallel and Distributed Processing” of Information Sciences, vol. 106, no. 3/4, pp. 237–258, May 1998. (Also appeared in International conference on Parallel and Distributed Processing Techniques and Applications, Sunnyvale, CA, August 9–11, 1996, 1431–1442)Google Scholar
- 10.J. L. Trahan, A. G. Bourgeois, Y. Pan, and R. Vaidyanathan, “Optimally scaling permutation routing on reconfigurable linear arrays with optical buses,” Proc. of the Second Merged IEEE Symposium IPPS/SPDP’ 99, San Juan, Puerto Rico, pp. 233–237, April 12–16, 1999.Google Scholar
- 11.H. Yamada, “Complete Euclidean distance transformation by parallel operation,” Proc. 7th International Conference on Pattern Recognition, pp. 69–71, 1984.Google Scholar