1996 | ReviewPaper | Chapter
Parallel distributed relaxation labeling
Author : S. Z. Li
Published in: Recent Developments in Computer Vision
Publisher: Springer Berlin Heidelberg
Included in: Professional Book Archive
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
A parallel distributed relaxation labeling (RL) method, called the Lagrange-Hopfield (LH) method, is presented. RL is treated as a constrained optimization problem. The LH method solves the problem using the augmented Lagrangian multiplier technique and the graded Hopfield network. The LH method effectively overcomes instabilities that are inherent in the penalty method (e.g. Hopfield network) or the Lagrange multiplier method in constrained optimization. Due to the use of Lagrangian multipliers, the normalization operation in traditional RL methods is dispensed with. This makes the LH algorithm fully parallel and distributed and is suitable for analog implementation. Experiments also show that the method is able to produce good solutions in terms of the optimized objective values.