Research On Visual And Infrared Image Fusion Method Under Low Illumination

Due to the obvious differences and complementary characteristics of visible and infrared images under the condition of low illumination,the visible image is easy to be influenced by the illumination,and the edge details are relatively rich.But the contrast and signal-to-noise ratio are reduced under the condition of low illumination,which makes it difficult to accurately determine the image target information.While the infrared image is a thermal image,which is not affected by illumination but belongs to the low-frequency image,so the edge detail information is fuzzy and low contrast,which is not suitable for target classification.According to the respective advantages of the visible and infrared images in low illumination,the two images are fused so that the fused new image can provide more target and background detail information.Aiming at the problems of the current image fusion algorithms,such as insufficient preservation of image details,the blurred target information and the poor self-adaptation of the algorithms,this paper proposes two fusion methods of visible and infrared images under low illumination.The main research contents are as follows:(1)Due to the poor visibility of visible images in low-light environment,an image fusion algorithm based on contrast enhancement and cauchy fuzzy function is proposed to improve the fusion effect of infrared and low-light-level visible images.The visibility of dark region of low-light-level visible image is improved by the adaptive enhancement of improved guided filtering,the intuitive fuzzy sets were used to construct the cauchy membership function and adaptive dual-channel spiking cortical model to fuse the lowfrequency and high-frequency components.Experimental results show that compared with other fusion algorithms,the algorithm can effectively enhance the dark area of the low-lightlevel visible image and retain more background information,thus improving the contrast and clarity of the fusion image.(2)In order to make full use of the information extracted from the middle layer and prevent information from losing excessively,a new image fusion network structure based on convolutional auto-encoder and residual block is proposed,which is composed of an encoder,a fusion layer and a decoder.First,the residual network is introduced into the encoder,and the visible and infrared images are fed into the encoder,the convolution layer and the residual block are used to obtain the feature map of the image.Then,the obtained feature map is fused by using an improved fusion strategy based on L1-norm similarity,which is integrated into a feature map containing the salient features of the source image.Finally,the loss function is redesigned and the decoder is used to reconstruct the fused image.The experimental results show that compared with other fusion methods,the method effectively extracts and preserves the deep information of the source image,which makes the fusion result have certain advantages in both subjective and objective evaluation.