3 February 2021 in IEEE Access
Image fusion is a visual enhancement technique that combines source images from different sensors to produce a more robust and informative fused image for subsequent processing or decision making. Infrared and visible light images share complementary properties that enable the production of robust and informative fused images. This paper proposed an infrared and visible image fusion method that improved the tetrolet framework to improve infrared and visible image fusion quality. First, the source image is enhanced by bicubic interpolation. The improved tetrolet transform then decomposes the enhanced source image; the high-frequency components are fused by convolutional sparse representation theory and combined with corresponding rules, and the low-frequency components are fused by defining ISER descriptors. Finally, we use the inverse transform to reconstruct the fused image. Qualitative and quantitative experimental results on five groups of typical infrared and visible image datasets demonstrate the proposed method’s effectiveness. The proposed method exhibits better performances on subjective vision and objective indexes compared with the other state-of-the-art methods.
26 December 2019 in IEEE Access
The purpose of image fusion is to merge a substantial amount of information, such as contour, texture and intensity distribution information from original images, with a fusion image. To retain a considerable amount of the fusion image while retaining the texture details of the source image and maintaining the edge of the source image, this paper proposes an improved infrared and visible image fusion algorithm that is based on total variation. First, source infrared and visible light images and a difference image were decomposed by a total variation model, and their respective cartoon and texture components were acquired. A fitness function was solved according to the entropy, standard deviation and edge similarity of the infrared and visible light images. The optimal combination weight of various kinds of texture and cartoon components was sought via a wolf pack intelligence optimization algorithm to acquire the final fusion results with high-quality contrast ratio and edge details. The experimental results indicate that the proposed method not only can preserve edge contour information about the original image but also can effectively retain its texture detail information. The method is superior to the traditional multiscale and sparse representation fusion method with regard to various indicators, such as subjective visual effect, mutual information, gradient information, structural similarity and visual sensitivity.