Research On Target Recognition System Based On Fourier Spectrum Polarization Imaging Techniques

In modern warfare, whether tanks, armored vehicles, or vehicles, artillery andother targets, an important prerequisite for the strategic plan is the ability toeffectively hide, not to be enemy found. Therefore, in order to try to increasetheir ability to survive in the wild environment, almost all of the battle targetshave various forms of camouflage. Correspondingly, a wide range for fast andefficient target identification system is studied by countries.In the wild environment, goals are often carried out by applying camouflagecamouflage or obstruction, and the purpose is to reach for goal hidden. In orderto achieve under various environmental, camouflage target can be identifiedeffective by the system. The target recognition system was designed based onstatic Fourier spectrum imaging techniques and polarization imaging technology.It’s fused by the spectrum target recognition image and the polarized targetrecognition image, at the same times, optimization algorithm for targetrecognition was proposed.Target recognition system was studied based on static Fourier spectrum imaging techniques, through analysis of the target spectral characteristics, andthe overall design of spectrum target recognition system was proposed.On thisbasis, spectrum imaging system was set up. Static Fourier transforminterferometer was designed, and the optical receiving antenna was designed.The spectrum distribution data of the target was acquired by the spectrum targetrecognition system.Based on the data of polarization, target recognition imaging system wasstudied. Theoretical basis for the existence of the target polarizationcharacteristics were analyzed. Function for target with a highpolarization-maintaining properties and the background was derived. Targetrecognition detection program was designed based on polarizationimaging.Polarization image target recognition system was set up bymulti-polarizer. Optical structure for suitable polarizing analyzer was selected,and the optical system for collecting the polarized light was selected. Polarizedimage data for target was acquired by the polarizing target recognition system.On the basis of the combination of two technologies, spectropolarimetric targetrecognition imaging system is constructed.Spectrum imaging algorithm was researched. Target feature information wasextracted used multi-characteristic wavelengths. In the analysis of a variety ofcharacteristic wavelengths extraction algorithm, principal component analysisand artificial neural network algorithm was combined with, and it constituted ajoint algorithm. It not only has the ability to quickly extract information characteristic wavelength, but also has the adaptability for an adaptive,self-organizing environmental. And as a basis, characteristic wavelengths wereextracted for spectrum of echo data. After experimental verification, it caneffectively obtain the appropriate characteristic wavelength groups by thismethod.The polarization imaging algorithms was researched. On the basis ofcomparing the polarization image for a variety of target recognition algorithm,according to the design requirements of the system, SSDA method of grayfeature type polarized target recognition algorithm was selected. The form ofpolarization image of the object recognition algorithm was similar to spectrumimage of the object recognition algorithm. Therefore, it was beneficial to the twoimage data fusion for the two systems. Polarization imaging algorithms andspectrum imaging algorithms were integrated. The two sets of target recognitionimage were fusion, and finally the spectropolarimetric fused image was got.Through comprehensive experiments can be seen, spectrum target recognitionimaging techniques or polarization target identification imaging technology isused alone, and there is no clear image object recognition problem. When theenvironment or target different backgrounds sometimes disguised way caused, iteven does not recognize. When image fusion technology was used, even avariety of environmental conditions and objectives camouflage mode is changed,it is still to be identified by the system.