Study On Signal Processing Technology Of IRFPA Imaging System

The signal processing technology of IRFPA imaging systems is one of the rapidlydeveloping and dual-use technologies. It has a lot of applications in many fields such astarget tracking, automaton navigation, assistant driving, etc.The design and realization of the image signal processing system based on staring IRfocal plane array around the application of Infrared focal plane imaging system areproposed. During the processing of IR imaging guidance, some IR image processingtechniques such as non-uniformity correction、noise restraint、enhancement of IR imageand real-time image enlargement、television display and ect have been deeply researchedand analyzed. Especially by using of the structure of FPGA+DSP in system hardwaredesign and taking full advantage of good performance of FPGA processor includereal-time operation、rebuilding design、low power consumption、flexible design、convenient debugging in system and ect. The design and realization of algorithms ofimaging processing such as two-points non-uniformity correction、expanded gray lineartransform、middle filter、bilinear interpolation enlargement and PAL/D standard televisionsystem are proposed.This paper mainly studies the relevant technology of design and realization ofinfrared focal plane imaging system. The main achievements of this paper are as follows:The techniques of IRFPA imaging system are discussed and studied principally. Areal-time IRFPA imaging system based on DSP+FPGA is designed and realized. It worksstably and can detect and track some typical targets both on ground and in sea.Put forward some IR image preprocessing algorithms by analyzing thecharacteristics of IR image include low contrast、non-uniformity、wide gray dynamicrange and describe the design and realization of some preprocessing algorithms based onFPGA in detail such as two points non-uniformity correction、middle filter、expanded graylinear transform and so on.The application of filter and prediction in spot target tracking is researched. Based onthe analysis of the error model of spot target location and displacement measurement, afilter prediction based tracking method for sport target is proposed.