Infrared and colorful visual light image fusion method based on color transfer and entropy information

Abstract

The invention discloses a blending method of infrared and multi-colored visible light images based on the information of multi-colored transmission and entropy. The process of the method is as follows: three channel images of R, G, and B of the multi-colored visible light images are calculated to obtain a typical value, thus obtaining visible light images with gray scale; the visible light images with gray scale and infrared images are decomposed by adopting non sampling Contourlet conversion; low frequency sub-band coefficient blending rules are constructed based on the infrared images and visible light physical characteristics, bandpass direction sub-band coefficient blending rules are constructed based on the combination of the entropy of local region direction information and region energy, the coefficient of transformation of source images are combined, and the coefficient of transformation combined carries out the non sampling Contourlet conversion to obtain blending image with gray scale; the multi-colored information of the visible light images is transmitted to the blending images by adopting a multi-colored transmission method based on 1 alpha beta color space, thus obtaining the multi-colored blending images. The blending method not only can effectively extract the abundant background information in the visible light images and the target information in the infrared images, but also can keep nature multi-colored information in the visible light images.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an image fusion method, which can be used to fuse infrared and color visible light images. Background technique [0002] Due to the different imaging mechanisms of infrared and visible light imaging systems, there is a great complementarity between infrared images and visible light images obtained in different environments. The visible light imaging system detects the target by absorbing the electromagnetic wave in the visible light band reflected by the target, while the infrared imaging technology mainly relies on the target itself or reflected thermal radiation for imaging. Visible light imaging systems are able to provide images with high spatial resolution, which can provide rich background information about the scene. However, visible light imaging systems are easily affected by weather conditions, and are generally used in situations with good light conditions. The infrar...

AI Technical Summary

Problems solved by technology

[0006] (2). The two-dimensional separable wavelet generated by the one-dimensional wavelet tensor cannot achieve sparse representation of straight lines and curves in the image, and the two-dimensional separable wavelet base is isotropic, and cannot be accurately expressed Therefore, the image fusion method based on wavelet transform is easy to introduce "artificial" effect, which reduces the spatial quality of the fused image to a certain extent.

[0007] (3). At present, most image fusion methods based on wavelet transform are based on the assumption that the source image is not disturbed by any noise when formulating the fusion rules of each frequency sub-band coefficient.

However, in some occasions, sensor images, especially infrared images, are inevitably affected by various noises

Therefore, most current fusion methods are highly sensitive to noise, and it is easy to mistake noise as useful information and transmit it to the fusion image, which affects the fusion performance.

Images