Image Matching Approach Based on Fuzzy and Genetic Algorithm

摘要

A novel image matching method is proposed to employ fuzzy transformation and fuzzy enhancement to preprocess the matching image pairs, and the genetic algorithm (GA) based on the real-valued coding, Latin square initialization, heuristic crossover and non-uniform mutation operators is used to resolve the matching parameters based on the affine transformation model. The simulation experiment and comparisons with other image matching methods are made to prove that the presented approach can solve the image matching problem with less matching time, higher matching probability and better matching precision than general image matching methods.