The Method And Realization Of Adaptive Infrared Image Correction Based On Domestic Detectors

Infrared detection and imaging is an indispensable means of detection in military fields such as infrared guidance,armed reconnaissance,target search and tracking.However,China's military infrared imaging has a large presence in developed countries such as Europe and the United States in terms of detection sensitivity and imaging stability.There are two main reasons for the gap: 1 the domestic detector has a low yield,the output response between the pixels is large and the stability is poor,and the non-uniformity is easy to drift with time,while the highperformance infrared detector has been embargoed by the US and Europe;Infrared image nonuniformity correction algorithms have their own limitations,and it is impossible to adaptively change non-uniformity correction parameters in complex scenes.At present,there are many algorithms for reducing image non-uniformity,and each has its own advantages and disadvantages.Designing an algorithm that can concentrate the advantages of various algorithms as much as possible and can be implemented on hardware is the focus of this work.At the same time,the system design is also critical.The low-noise system circuit can improve the signal-to-noise ratio of the output image signal of the detector.The excellent software architecture protects the infrared image processing.To this end,the paper carried out the following three aspects of research:1.Taking the domestic detector as the research object,research the detector data manual,select the core chip according to the characteristics of the required driving voltage,driving signal and output image of the detector,and thus establish the hardware architecture consisiting of power board,FPGA board,driver board and front-end board.Completes the schematic drawing of four low-noise boards and debugs the board.2.Introduced Qsys platform,Avalon interface protocol,Nios II processor and ?C/OS II system.With Qsys as the platform,the detachable system software architecture was established through Avalon bus protocol.The software and hardware collaborative design based on FPGA and Nios II,written the driver and by the OCC on-chip non_uniformity correction,allowing the detector to output an image signal with a relatively low non-uniformity.3.The causes of the non-uniformity of the infrared system are analyzed.The principles and characteristics of several popular image reduction algorithms are introduced and analyzed.An algorithm based on bidirectional registration non-uniformity correction is proposed.Through the simulation to realize the function and transplant into the hardware system,while reducing the non-uniformity of the image,it has a faster convergence speed and can suppress the "ghosting" phenomenon,so that the infrared imaging system can adaptively adjust the correction parameters according to the scene.The imaging quality of the infrared imaging system is improved,and the long-term stable operation of the infrared thermal imaging system based on the domestic detector is ensured.Based on the research of these three aspects,this paper proposes an algorithm based on bidirectional registration non-uniformity correction and applies it to the designed system architecture.It realizes adaptive infrared image correction based on domestic detectors and overcomes various disadvantages of current image processing algorithms.So that domestic detectors can also output high-quality images,reaching the advanced level in foreign countries.