一种用于目标探测的图像多层级融合和增强方法

In this paper, a novel infrared-to-visible image fusion algorithm for enhancing contrast and visibility is proposed. A multi-level method based on the characteristics of images and the properties of the targets is designed to complete the image fusion process, where a contrast enhancement method is added in the low-frequency information of the layered images and the edge information is enhanced in the high-frequency information using the correlation between the low- and high-frequency components. In the experiments, three groups of infrared-to-visible images were used to demonstrate the effectiveness of the multi-level fusion method. All the evaluation indexes, such as standard deviation and information entropy, were significantly higher than other existing methods. Thus, the experiental results verified the effectiveness of the proposed image fusion methods. The quality of the fusion images was improved for better differentiability in terms of contrast and features of the targets