Research On The Key Techniques Of Uncooled Infrared Thermal Imaging System

In recent years, infrared imaging technology has been developed rapidly. It playsan important role in the military and commercial fields. Compared with thetraditional cooled thermal imaging systems, uncooled infrared imaging systems havethe advantages of low price, small size, light weight and low power consumption andconsequently gain great attention in the development of infrared imaging systems. Inthis dissertation, we review the fundamentals, analyse the determining factors andtheir effects on the quality of an infrared imaging system and study the main featuresof the generated images. Different approaches are considered for correcting thenonuniformity of infrared focal plane arrays, compressing infrared images with highdynamic range and enhancing image details.The image quality of infrared systems can be severely afftected by thenonuniformity of infrared focal plane arrays. On the analysis of the nonuniformityand its causes, both black-body calibration and scene-based nonuniformitycorrection methods are discussed in this paper. Aimed at the ghosting anddegradation which arise in traditional scene-based correction methods, a novelmethod based on constrained Gaussian filtering is proposed here. Experiments showthat this method can increase the signal-to-noise ratio (PSNR) of infrared imagesfrom19.43dB to30.71dB. The proposed method can obtain better correction results and effectively suppress the ghosting and scene degradation at the same time.It is very common that infrared images have low contrast and low SNR.Therefore it is necessary to enhance them. In this paper, the main techniques ofdynamic range compression as well as the enhancement of image details arereviewed. Since conventional algorithms are incapable of preserving sufficientinformation of the original image, we propose a new method based on the optimalmapping curve and apply it to the compression of dynamic range and enhancementof infrared images. The experiment carried out on the basis of a large amount ofinfrared images show that compared to the existing method, the proposed methodcan minimize the mean square error of the enhanced image compared to its origin.Meanwhile, it can significantly improve the overall visual effect of the image. Thisalgorithm is fast, highly automotive and suitable for real-time processing of infraredvideo.In order to implement the real-time processing of the signals from infraredimaging systems, a FPGA-based signal processing system is designed. This systemtakes FPGA as its core and it achieves all the functions including the drive of thefocal plane array, nonuniformity correction, dynamic range compression, detailenhancement, output of analog video, output of digital signals, and electronic zoom.Finally, for the developed infrared thermal imaging system, a detailedassessment involving the power consumption, nonuniformity, noise equivalenttemperature difference, minimum resolvable temperature difference, imaging timeand system response time is carried out based on the existing test platforms andimaging processing software like Matlab.The research in this paper has great significance and practical value on theuncooled thermal imager successfully developed by Chinese Academy of Sciences.