Polarization Image Acquisition And Polarization Feature Fusion Based On FPGA

Imaging detection of targets in complex scenes,especially the identification of false,disguised,and stealth man-made targets has become an important research topic in today's battlefield environment analysis,intelligence acquisition,and target reconnaissance.Traditional intensity imaging detection only uses the sensor to capture the light intensity information of the target.In practical applications,the light intensity information will be weakened to a certain extent when the measured target radiation passes through a complex background environment,resulting in blurred images,low accuracy,and recognition poor ability.The polarization imaging technology can not only obtain the light intensity information and polarization information of the measured target,but also highlight the artificial targets that are hidden or indistinguishable from the naked eye,thereby improving the ability of intelligence acquisition on the battlefield.In this paper,based on the polarization optical imaging technology,the research of FPGA based polarization image acquisition and polarization feature fusion is carried out.Firstly,the basic theory of polarization imaging detection is analyzed,and the polarization image acquisition scheme is designed by combining the polarization optical characteristics and the target imaging principle;secondly,the hardware and software design of the polarization image acquisition system is implemented with FPGA as the main controller,including polarizer calibration,system circuit Design,PCB design welding,driver design,host computer writing,etc.,and complete the construction of the experimental platform;then register the obtained polarized image,and apply the image fusion method of discrete wavelet transform and discrete Curvelet transform to realize the polarization image Integration with light intensity images;finally,under different experimental scenarios and experimental conditions,a large amount of polarized image acquisition and processing work was completed,and the objective evaluation and evaluation of image fusion quality were made by indicators such as edge intensity,average gradient,information entropy,contrast,etc.analysis.Experimental data shows that the method proposed in this paper can obtain polarizedimage data at high speed,can be saved in JPG image format and AVI video format,and can detect real-time targets and dynamic targets in real time.Abundant and other advantages,can quickly and accurately achieve the target detection and recognition.