A Qulity Evaluation Method Of Image Fusion Based On IGM Algorithm

With the rapid development of information technology and digital images technology have become more and more integrated into our life. While the digital image would face with quality loss and multi-source image analysis demand when apply and transmit. Image fusion technology can solve the above problems. Image fusion is divided into two situations: one of which is images collected from multi-source channel on a same target, the other one is a single source multiple imaging set’s fusion images’ imaging situation and times are different. Image fusion refers to maximize extract useful information from images and upgrade image quality by image processing technologies and computer technologies Researchers proposed a lot of different fusion algorithms, which have good performance in different fields, but the final users and judges of fusion images are human beings, and according to different needs, the image fusion quality evaluate method is different.According to the difference of evaluation object, image fusion quality evaluation method is generally separated from two types: subjective evaluation method and objective evaluation method, the subjective evaluation method refers to select a set of evaluation group which contains a number of experts; fused images by different fusion methods are submitted to evaluation group. The evaluation group should give a score based on their own judgment. Finally rank the score according to the comprehensive score which is given by evaluation group. While the subjective evaluation method has its shortcomings, first of all, the fusion quality evaluation results will be influenced by objective factors affect larger, such as image display medium, image viewing angle, light and other factors. In addition, the evaluation accuracy of fused image is not very high which obtained by subjective evaluation method, especially the identification of subtle difference between fused images. In addition the cost of subjective evaluation method is very high, and the efficiency is very low, all of these restrict the use of subjective evaluation methodsSo people try to find an objective quality evaluation method on fused images, which can completely solve the above problems. The current research in this fields can separated from some methods below according to the different reference index: methods based on statistical analysis, methods based on information and methods based on human visual system. In this paper we analyze the implementation and comparison of different kinds of fusion image quality evaluation methods. We proposed a new fusion image quality evaluation method based on Internal Generative Mechanism—PIGMQAThe main research contents of this paper are as follows:1. Introduction and analysis of the image fusion quality evaluation research status and the main fusion quality evaluation method. 2. According to the research on IGM of human visual system explore the mechanism of the human visual system in internal research and explore the process of HVS on image similarity evaluation and image quality evaluation model. Based on the above research, we put forward an objective image fusion quality evaluation method---PIGMQA, through a series of experiments, verified the effectiveness of PIGMQA from multi-dimensions. 3. Further, we improved the PIGMQA image fusion quality evaluation algorithm, discussed the feasibility of PIGMQA on universality image fusion evaluation method and put forward the corresponding improvement method.The PIGMQA image fusion quality algorithm proposed in this paper has a great improvement on the accuracy of fusion image quality evaluation, and it is more close to the subjective image quality evaluation methods’ results. This method is of great importance on the process of different situations, image fusion, image restoration quality evaluation and the selection on fusion method.