基于智能型红外焦平面电路液氮成像系统的研究与算法实现  

Research and algorithm implementation of liquid nitrogenimaging system based on intelligent infrared focal plane circuit

作　　者：陈相洪 黄文刚[1] 黄晓宗[1] 陈彦冠 史凡萍 CHEN Xiang-hong;HUANG Wen-gang;HUANG Xiao-zong;CHEN Yan-guan;SHI Fan-ping(The 24th Research Institute of CETC,Chongqing 400060,China;The 11th Research Institute of CETC,Beijing 100015,China)

机构地区：[1]中国电子科技集团公司第二十四研究所,重庆400060 [2]中国电子科技集团公司第十一研究所,北京100015

出　　处：《激光与红外》2023年第8期1203-1212,共10页Laser & Infrared

基　　金：模拟集成电路国家级重点实验室基金项目(No.6142802190101)资助。

摘　　要：提出了一种智能型红外焦平面电路液氮成像系统,系统采用UART转SPI数据传输,上位机VB软件开发SPI配置界面,可提高SPI寄存器批量配置效率和准确率,减少外界干扰产生误码和乱码情况,并结合FPGA、自研的IRCamExpert-CETC图像处理软件和图像采集卡搭建软硬件成像系统,本系统具有红外视频图像在线监测和多芯测试能力。本文进一步提出了基于单片像元级ADC的背景减除、盲元补偿、非均匀性校正、开窗、像素合并、空间滤波、图像均衡变换、时间延迟积分多种图像算法和总体算法结构,并通过对一款640×512阵列的智能型红外焦平面电路进行测试,验证了单片红外焦平面多功能图像算法,并实现了系统对智能型红外焦平面全参数的性能评估。采用像元级进行图像处理的方式,降低了采用算法级进行后续图像处理的难度,减少了图像处理算法消耗的资源。多功能图像算法为智能型红外焦平面的工程化应用提供了新技术途径。In this paper,an intelligent infrared focal plane circuit liquid nitrogen imaging system is proposed,which uses UART to SPI data transmission,upper computer VB software to develop SPI configuration interface,to improve the SPI register batch configuration efficiency and accuracy,and reduce the condition of external interference generated error and disorderly code.The hardware and software imaging system are built by combining with FPGA,self-developed IRcamExpert-CETC image processing software and image acquisition card.This test system has infrared video image online monitoring and multi-core test capabilities.Furthermore,the multiple image algorithm and overall algorithm structure is proposed based on monolithic pixel level ADC including background subtraction,blind pixels compensation,nonuniformity correction,windowing,pixels merging,spatial filtering,image equilibrium conversion,time delay integral.Based on a 640×512 of intelligent infrared focal plane array circuit test,the single-chip infrared focal plane multi-function image algorithm is verified to realize the performance evaluation of the whole parameters of the intelligent infrared focal plane.The pixel level image processing method is used to reduce the difficulty of subsequent image processing at the algorithm level and reduce the resources consumed by image processing algorithms.The multifunctional image algorithm provides a new technical approach for the engineering applications of intelligent infrared focal plane arrays.

关 键 词：智能型红外焦平面 液氮成像系统 多功能图像算法 全参数性能评估 

分 类 号：TN219[电子电信—物理电子学] TP391[自动化与计算机技术—计算机应用技术]