Ambiguity reduction by relaxation labeling
作者：
Highlights：
•
摘要
A new relaxation labeling algorithm for a general n-node/m-label system is developed and its theoretical as well as application aspects are analyzed. The convergence properties of the new relaxation algorithm are presented. Physical interpretations of the mathematical terms are discussed. Finally, an experiment of applying the new relaxation labeling algorithm to perform an object identification task in an intelligent robotic system with two cameras, is described.
论文关键词：Information integration,Neural network,Relaxation labeling
论文评审过程：Received 11 February 1993, Revised 6 July 1993, Accepted 6 August 1993, Available online 19 May 2003.
论文官网地址：https://doi.org/10.1016/0031-3203(94)90025-6
