Infrared Image Acquisition And Processing System Based On FPGA

In recent years,with the continuous development of infrared imaging technology,its applications have rapidly expanded to various fields such as military,aerospace,medical,and communications.However,due to the influence of the infrared detector device itself and the application environment,the original infrared image is collected by the infrared imaging system often exhibits problems such as poor resolution,low contrast,high noise,and blurred visual effects.Therefore,it is necessary to deal with the collected original infrared image relevant processing can be satisfied with the requirements of various fields.This article applies high-resolution IRFPA to build an image acquisition hardware circuit system and collects infrared images of the required scene,then exploits software platforms such as MATLAB and Model Sim to model and simulate the processing algorithms of the collected images,and finally uses FPGA to realize infrared images treatment.First of all,this paper describes the research background and development status of infrared imaging systems,and clarifies the necessity of processing and optimizing the collected original infrared images.At the same time,theoretical analysis and experimental comparison of traditional processing algorithms are carried out for the blind pixels and non-uniform noise that are common in infrared images.In addition,the commonly used infrared image quality evaluation methods are introduced.Secondly,this paper ameliorates the traditional blind pixel compensation algorithm to enable it to detect blind pixels in real time,thus avoiding the new situation that blind pixels cannot be compensated,and introduces the correlation and interpolation point calculation method in the AMMC algorithm for blind pixels.The correlation between the compensated pixels and the surrounding pixels is emphasized.Then the two generalpurpose single-frame infrared image non-uniformity correction algorithms and classic image denoising algorithms are experimentally compared,and a processing method suitable for this system is detected.Secondly,an adaptive platform histogram equalization method is used to improve the contrast of infrared images and finally the rainbow coding method is used to process the pseudo-color of the infrared image,which enhances the human eye's ability to distinguish infrared images.The experimental results show that the improved blind pixel compensation algorithm in this paper can realize real-time detection.Compared with the traditional median filter algorithm,the correlation is increased by3.56%,and the non-uniformity of the processed image is reduced by 0.2% compared with the original image.The mean square error an increase of 20.6%,although the nonuniformity decline is limited,the visual effect is favorable.Finally,this paper optimizes the infrared image acquisition hardware circuit to initially improve the image contrast at the hardware level,and also improves the implementation method of the blind pixel compensation module,which reduces the resource utilization rate of the module by 70%,and the key modules of image processing algorithm are simulated and verified,and the interface signals and implementation methods of the simulation modules are given.