Design And Implementation Of Image Fusion Correction Algorithm Based On FPGA

Image fusion is to fuse the different band information of the same target image according to some criterion,so that the fused image has a better character than any one of the fused images,so as to reflect the actual information more accurately.Taking the infrared image and the micro optical image fusion system as an example,the dual channel optical axis of the actual image fusion system is not strictly parallel because of the assembly reasons,so it will cause a certain offset between the images produced by different channels.And because of the distortion of the image itself,the image fusion operation is often required after the distortion,rotation,translation and or zooming correction of the image.The traditional correction method for the mismatch is to correct the distortions,translations,rotation and scaling operations.Every step on the processor needs to take up the corresponding chip resources and need a relatively long operation time.This thesis focuses on the research of image fusion registration based on FPGA.This paper first introduces the development status and basic theory of image registration in the field of image fusion,and the basic principles involved in the research,and analyzes the distortion correction,translation correction and rotation correction from the basic principles.A set of fast image registration algorithm based on FPGA is derived from the mathematical basis and correction method of positive and zoom correction.Based on the distortion parameters of the infrared image itself and the translation parameters,rotation parameters and zoom parameters relative to the micro optical image,the algorithm is set up to correct the infrared image lookup table.The corresponding location of each pixel in the infrared image corrected image in the uncorrected infrared image is stored in the lookup table.Through this look-up table,the algorithm proposed in this paper can correct the four mismatch problems by single operation.It can save the chip resources of the processor in the image fusion system,reduce the power consumption and improve the operation speed.In this paper,on the basis of the algorithm,the infrared and micro optical image fusion correction system based on FPGA is designed.The modules of the hardware are briefly introduced.Finally,the simulation tests of each module are carried out,and the results of the test are analyzed objectively.