Research On Real - Time Signal Processing Technology Of Infrared And Micro - Light Fusion

Infrared and LLL(Low Light Level)/visible image fusion technology makes full use of the advantages and complementarities of two kinds of imaging techniques, which has become a hot research topic in the field of image fusion technology, and has been widely used in many aspects including military and civilian. The requirement of the system performance is promoted with the broadening of the application field of the image fusion system. There are lots of issues required solutions, such as how to improve the convenient and real-time performance, how to get lower cost and power consumption, and how to apply some good algorithms in the system. To solve these issues above, this paper makes a research of image fusion algorithm and hardware implementation based on the infrared and LLL image fusion night vision system.Firstly, the development of fusion algorithms and systems domestic and abroad are introduced in this paper, then the design method of normal fusion system and the implementa-tion of this system including hardware structure and software design are explained in detail.Secondly, the preprocessing and registration of double source image are studied and used on FPGA eventually, which including contrast enhancement of infrared image and the recursive processing of the noise of LLL image. Use the affine transformation for image registration and bilinear interpolation algorithm for image scale.Thirdly, according to the computing power and parallel processing characteristics of FPGA, the traditional image fusion algorithms are studied and then improved or optimized, and achieve Laplacian pyramid image fusion and color transfer false-color fusion algorithm in the system at last. A pyramidal decomposition and reconstruction method based on the same resolution is proposed for changing the traditional pyramid, and an optimal method to lower the high computational complexity of color transfer, so that the work of calculation is greatly reduced. The result of fusion image is good and meets the demand of real-time performance.Finally, in order to solve the problem of saving image in use process, propose a method based on JPEG compression. All the work of JPEG image compression was finished in FPGA. Experiments shows that after decompression, the compression coding data of the system is correct and the quality is good, the average compression ratio is about 15:1.