Research And Implementation Of Infrared Detector On-chip Calibration And Infrared Image Defect Detection

Infrared detector has a wide range of applications in the military and civilian fields.To further reduce the volume and power consumption,TEC-less uncooled infrared focal plane detector has been developed.To reduce the nonuniformity of the detector,the bias circuits which can have the function of on-chip nonuniformity correcton are integrated.For on-chip calibration has high real-time performance and obvious correction effect,in recent years,the newly launched traditional or TEC-less infrared detectors by the domestic and foreign infrared detector manufacturers have generally started to have onchip correcton functions.With the development of infrared detectors,there have been many new demands in the field of infrared detector testing.Generating on-chip calibration data quickly and accurately is very important for detector testing.In addition,infrared images may have some abnormal defects.Detecting these defects helps to understand the quality of the detector.In the subsequent applications,relevant image processing algorithms can be targeted for improving the imaging quality of the detector.Based on this reserach background,this paper makes a deep research on the on-chip calibration function and the infrared image defect detection.This paper has designed and implemented a set of infrared detector test system.Different from the laboratory's previous generation of test system,this test system uses FPGA to build an embedded system based on Nios II processor,which enables the test system to complete more and more complex functions and lay a good foundation for the further upgrade of the system.The test system can complete the functions of detector driving,detector output collection,data transmission based on Camera Link protocol and USB communication,which makes this paper has a good software and hardware foundation for on-chip calibration research.Based on the test system,this paper designs and implements a stable and fast onchip calibration method based on the dichotomy.The calibration process roughly takes only 0.3s,and the generated correction data reduces the fixed image noise by 92.6%.In order to further improve the effect of on-chip calibration,this paper also designs and implements a fine-tuning method for on-chip calibration effects,which further reduces the non-uniformity of the infrared focal plane array.Fixed image noise decreased by 24.3%.In this paper,the influence of on-chip calibration on some parameters of the detector has been tested and verified.On-chip calibration can effectively reduce the fixed image noise and average noise voltage of the detector,increase the dynamic range of the detector,but can not reduce the responsivity non-uniformity.Aiming at the detection of infrared image defects,this paper designs and implements an algorithm that can be implemented using Matlab software platform.The algorithm is mainly based on the principle of threshold segmentation,and can effectively detect the dark and bright defects in a total of 14 kinds proposed in this paper.The overall detection rate is about 94.1%.