Entropy-based image fusion with continuous genetic algorithm

This paper proposes a novel image fusion method that uses continuous Genetic Algorithm (CGA). Image fusion provides a single output image from a set of input images, captured from different viewing conditions, by different sensors, and at different times. Image fusion can be divided into three processing levels: pixel, feature and decision. Genetic algorithms are a nonlinear optimization technique that seeks the optimum solution by modifying the input vectors. In the pixel level image fusion (PLIF) process, the weights used for fusion need to be optimized to form a new composite image with extended information content. In this paper, we present a novel, entropy-based image fusion algorithm that uses CGA. Experimental results show that CGA based image fusion outperforms other point-rules based PLIF algorithms.