Research On The Technology Of High Sensitivity And High Spectral Image Information Processing

Electromagnetic spectrum LWIR(Long Wave Infrared, LWIR, 8 ~ 14μm) spectrum in physics may also be called thermal infrared spectrum. Most of the thermal radiation energy of objects on the Earth’s surface at room temperature(300K) is concentrated in the band. Surface temperature and size of object thermal radiation energy is directly related. People can receive radiation signals to calculate the temperature and emissivity. Thermal infrared detection Owing atmospheric window, penetration ability, and working all day and night, and can be widely used in civilian and military fields.But because high performance thermal infrared focal plane is difficult to develop, mature thermal infrared detector performance compared to short wave infrared detectors have a larger gap. Especially in the detection of weak infrared signals, imaging is still unable to reach a higher quality. Therefore, within the constraints of hardware conditions, how to further breakthroughs in the target small radiation detection limits, raise detection sensitivity in weak radiation of a thermal infrared field is an eternal theme.There have been some studies abroad in detection methods and processing techniques for higher sensitivity required, but mostly confined to the non-uniformity correction algorithm. These methods have certain limitations, either to enhance the performance of the temperature powerless, either to reduce spatial noise powerless. Document[71] through digital TDI technology have made system NETD16.4 level, but there are still many limitations. Most of the processing method only looks at the time noise equivalent temperature difference(NETD),but less focus on space noise which more influence on the final image of the object recognition capabilities. So it is necessary to consider temperature sensitivity of thermal infrared images from more perspectives.So this paper focuses on the remaining non-uniformity of thermal infrared image and a comprehensive analysis of the different non-uniformity correction algorithm. Study on the thermal infrared image processing method of the non-uniformity correction, in order to further reduce the remaining non-uniformity. Carry on research in the calibration factors to further improve the sensitivity of the final image. This paper proposes a new method to improve weak signal detection sensitivity in thermal infrared radiation with the existing hardware condition. This method can be made on the basis of the existing hardware, greatly improve the system of small thermal infrared radiation detection capability. This adds a viable means to detect the faint infrared signal.(1) System Noise ModelA comprehensive analysis of the principle of thermal infrared imaging system noise and causes is proposed. By separating the temporal noise and spatial noise, it noted that spatial noise is more important factors affecting the final sensitivity. It highlighted the reason of the existing algorithms causing the remaining non-uniformity. It pointed out that such a system error can not be removed by non-uniformity correction algorithm, and gives the theoretical size of remaining non-uniformity. It provides a basis for the proposed follow-up time and space sensitive superposition techniques. This work not only points out the lack of non-uniformity correction, and gives the remaining non-uniformity of the theoretical value. It provides a theoretical basis for the proposed follow-up time and space sensitive superposition techniques. Therefore, a detailed study and analysis of the residual inhomogeneity model has a certain innovation.(2) thermal infrared image non-uniformity removal methodCarry on research on the algorithm for the non-uniformity of system response characteristic. For lack of different non-uniformity correction algorithm, it firstly proposed two processing method after non-uniformity correction. Respectively, for striped noise and fixed background noise, in order to further reduce non-uniformity of thermal infrared staring image, paper provides a new way. Also studied the calibration process of various different factors, we can come to get the best result of non-uniformity correction calibration method. Through the experiment, the effectiveness of the method is verified. This is firstly proposed non-uniformity correction processing method and calibration methods. It has a strong engineering application value, innovative.(3) time and space alternately stacking techniqueFor Thermal infrared detection for high sensitivity applications it proposed space-time alternately stacking technique. On the basis of the use of the foregoing method, based on a self-developed time NETD32 mk thermal infrared imaging system it achieved the ultimate space NETD3.9mk detection sensitivity, and achieved detection sensitivity to a weak signal far more than the detector. By combining laboratory and outdoor flight data to achieve high sensitivity thermal infrared detection, the effectiveness of the method is verified. This part of the work achieves a high sensitivity thermal infrared detection and obtains high-sensitivity thermal infrared images of the sea,for the first time, with strong innovation and engineering value.(4) thermal infrared hyperspectral image processing technologyStudy analysis of the signal component in thermal infrared hyperspectral imager. Combined non-uniformity correction processing method, based on 863 thermal infrared hyperspectral imager flight data, proposed a practical method for hyperspectral data processing. Finally paper get high-quality thermal infrared hyperspectral data sets and lay a good foundation for further application. It has a strong engineering application value, but also has some innovative.