Full-focused image fusion in the presence of noise

摘要

The implementation and analysis of the algorithm for the full-focused image fusion in the presence of noise are presented. Three methods of combining noisy images are considered: without pre-processing and post-processing, using prefiltration of original images, using post-filtering of the fused image. The database of test scenes created by the authors was used for testing the proposed algorithm for full-focused image fusion. Additive white Gaussian noise was considered as an noise model. Two-stage digital image processing scheme, based on principal components analysis was used as a filtering algorithm. Quantitative and visual results are shown and demonstrate the main features of the proposed algorithm.