Selection Method for Scene Matching Area Based on Information Entropy

ZHANG Xiao-chen,FU Meng-yin
DOI: https://doi.org/10.3969/j.issn.1001-506x.2011.05.26
2011-01-01
Abstract:A new selection method for the scene matching area(SMA)based on information entropy is proposed. To improve the reliability in selecting the SMA,the scene matching ability parameters including image entropy,Frieden entropy and normalized mutual information are defined based on entropy theory to evaluate the information content,stability and unique characteristic of the SMA.According to that,the best SMA is selected step by step.Experiment results validate that the SMA selected by this method has high matching probability and low matching errors.The method,which is logical and reasonable,can meet the matching requirements.
What problem does this paper attempt to address?