Research On Infrared And Visible Image Registration And Fusion Algorithms And FPGA-based System Development

Infrared imaging has a wide range of uses in military and civilian fields,and infrared and visible image fusion is one of the important research directions in the field of multi-source image fusion,especially the fusion of long-wave infrared images and low-light visible images.Long-wave infrared penetrates well and is able to overcome low illumination,rain,fog,occlusion and other environmental imaging with good image contrast and significant target.The low-light camera is able to image clearly in low-light environment with rich image details.By processing the two images with alignment and fusion algorithms,the advantages of both can be combined into one to improve the detection of scenes.The research and system development of binocular alignment and fusion algorithms for infrared and visible images are carried out in the paper,and the specific work includes:(1)The thesis proposes a binocular registration method for infrared and visible based on the feature points of markers.Using infrared and visible fusion system to shoot the self-made dual-band calibration board image,analyze the binocular image to establish a parallax model,use the feature points on the marker board to solve multiple registration matrices at different distances,and make a look-up table.When observing the target,the registration matrix is searched according to the working distance to realize the registration of infrared and visible light images.This method improves the shortcomings of slow registration speed and large errors caused by the large consistency difference between infrared and visible image characteristics.The experimental results show that,according to the observation distance to find the transformation matrix for registration,the effective registration can be achieved within the range of 5m?500m.(2)The thesis proposes an infrared and visible image fusion algorithm based on side window filtering technology.The algorithm uses side window filtering(SWF)technology to decompose the source image into a base layer and a detail layer,which effectively preserves the edges of the source image and reduces artifacts and halos near the edges.Then,the base layer image is normalized and non-linearly processed,and regularization parameters are introduced to adjust the ratio of image fusion,control infrared energy and image contrast,and improve the robustness of the fusion algorithm.In addition,in order to make the fused image have more visual details and less noise,side window guided filtering(SWGF)is used to optimize the detail layer image fusion.The experimental results show that the fusion image of this method has rich details and prominent targets,and both subjective and objective evaluation are better than the existing 9 methods.(3)The hardware platform of infrared and visible light fusion system is built.The image registration algorithm and linear fusion algorithm proposed in this paper are transplanted to the hardware platform.The FPGA is used as the core processor,and the peripheral arm controller,video acquisition and output,synchronous focus,power supply and communication circuit are used to form the fusion system.The output of fusion image is realized by running on hardware..