Research On Multi-focused Image Fusion

The acquisition of the same accurate description of scene images, image fusion [this description still don't feel good, you can refer to others, such as multi focus image fusion is an important branch of image fusion, is widely applied in machine vision, target recognition, digital cameras and other fields. This technique is intended to extracted from the image in the same scene with different focus areas into all the objects are clear. In practical application, some optical lenses often fail to make the same scene of multiple targets in the same region, in this case the resulting image is not comprehensive and clear. This creates a fusion a comprehensive and clear image from multiple images. It is of great significance for studying multi focus image fusion. This paper mainly in multi focus image fusion based on the existing research results Do further research on the basis.This paper focuses on the research of multi focus image fusion technology, and the specific research work is as follows:1.In order to overcome the existing literature in a single dimension image fusion for inaccurate identification of non focused regions leading to fusion effect is poor, this paper for fusion results to measure the scale of the problem presents multiscale image fusion algorithm, in order to overcome the problems existing in the literature single fusion scale brought about, such as determining the focus area error, caused the fusion effect is poor.2.To overcome the existing literature for non connected unfocused region recognition inaccurate proposed an algorithm can have better fusion effect in the field, good improved available in the literature for non connected unfocused region judgment error shortcomings.3. In order to overcome the Gibbs effect and improve the edge separation degree and peak signal to noise ratio test index, this paper proposes a bilateral filter based image fusion algorithm and genetic algorithm optimization of experimental parameters at the edge of the separation degree and information entropy and peak signal to noise ratio index compared with the existing literature has a great advantage.At last, the author summarizes the work of this paper, and makes a summary of the difficulties in the course of the study, and makes a prospect for the future research.