Research On Thermal Infrared Image Enhancement And Its Fusion With Visible Image Fusion And Application

The thermal infrared image technology uses the infrared thermal imager to obtain the thermal information of the objects,thus deduce the characteristics of the objects and their interaction with the environment,playing an increasingly important role in the scientific production.However,due to its imaging principle,the thermal infrared image is gray level,low contrast,visual blur,so that it has a great deal of inconvenience in the production application,to a certain extent,limited the further development of thermal infrared image technology.Therefore,it is necessary to study the enhancement technique of thermal infrared image.In this paper,the characteristics of gray image of thermal infrared image are studied,and the thermal infrared image is fused with visible light image to make up the low contrast and blurred background of the thermal infrared image by using the high resolution and richness of the visible image.Based on this,the main work of this paper is as follows:(1)The pseudo-color coding algorithm of thermal infrared image based on piecewise equalization is studied and designed.Thermal infrared images are gray-scale images,because the human eye can distinguish between the interval narrow,and vice versa for color image resolution is very sensitive.Therefore,in order to make the thermal infrared image more convenient to observe and apply,research and implementation of the thermal infrared image pseudo-color coding algorithm.Based on the distribution of the gray pixel values of the thermal infrared image,the algorithm is proposed to equalize the pixels so that the pixel values can be scaled to the range of [0 ~ 255] in the original ratio,and then the pseudo color transform is carried out by using the rainbow coding method.In this paper,the experimental results show that the images processed by the pseudo-color method designed have obvious advantages compared with several traditional methods,and the subjective observation effect is also improved.(2)Research and implement of thermal infrared/visible image fusion algorithm based on NSCT transform.The image is divided into high frequency sub-band and low frequency sub-band by NSCT.The low frequency sub-band contains most of the information of the image,and the high frequency sub-band contains the details of the edge and texture of the image.The fusion rules of the high frequency sub-band and the low frequency sub-band design are different,which can highlight the goal of fusing the image and keep the edge,texture and other details.Compared with other traditional fusion methods,the fusion effect has some advantages on the objective standard,the objective and background of the subjective experiment result are clear.The pseudocolor coding of the fusion image realized by this method is improved in both subjective and objective aspects.(3)Design and implement a thermal infrared/visible image fusion algorithm based on target area.The most important goal of fusion of thermal infrared images and visible light images is the ability to utilize the target information in the thermal infrared image and the background information of the visible light image.Therefore,if the target and the background were fusion processing,it will achieve a better fusion effect in the fusion process.Firstly,the time information and spatial information detection target of thermal infrared video are used to determine the target area.NSCT transformation is performed on the original image,the target region and the background region are designed with different fusion strategies to complete the image fusion.The experimental results show that the fusion image has a clear target and the background information is intact.The pseudo-color image obtained by fusing the coded image is the best.(4)Design and implement the thermal infrared image integrated application platform.The platform can capture the video by the thermal infrared / visible double spectrum camera or the local loading mode,then use the thermal infrared image processing method realized in this paper to detect the target of the loaded video,the fusion of the thermal infrared/visible video,and the fusion image to carry out pseudocolor coding,and playing the various stages of the video processing function.The experiments of the two sets of video,the effectiveness of the platform is proved.