Research On Low-contrast Target Detection And Tracking

Low contrast target such as space debris detection and tracking is significantly important for science research and military applications. This dissertation focuses on the critical theory and concentrates on engineering realization even verification to improve the low contrast space target detection capability and tracking stability, including target characteristics analysis, daytime sky background estimation, image signal reconstruction and target detection algorithm.Target detection mechanism and current development of low contrast target detection and tracking is discussed in detail at first. If target’s signal-to-noise ratio(SNR) is less than 3dB, it is defined as low contrast target. Image enhancement algorithms and background suppression methods of low contrast target detection are discussed; the means to separate target and background based on adaptive threshold and boundary extraction is analyzed; and low contrast target detection technology based on clustering and pipeline is also discussed.Target characteristics are analyzed from following aspects: the basic luminosity model of geometry structure and optical characteristics including size is established; target motion characteristics including angle velocity and angle acceleration is analyzed, and it could provide the basis for dynamic transfer function of target detection and tracking system. The spectral reflection of typical target is studied in depth, and an infrared characteristic extraction algorithm based on double wavelength band infrared spectrum is analyzed and presented. The effectiveness of the proposed algorithm is verified by the outfield experimental data.Atmosphere transmission and daylight background usually changes dynamically with strong uncertainty, and they have obvious effect on the performance of based-ground detecting and tracking system. The atmosphere absorption, scattering coefficient, atmosphere background radiation, transmission index, and sky background brightness are calculated, and the influence that the atmosphere inflicts on light absorption and light scattering is analyzed when it is distributed in vertical direction. Furthermore, the impacts that atmosphere light scattering inflicts on the contrast of target and atmosphere turbulence applies on target dispersion are analyzed in detail.The static and dynamic performance of low contrast target detection and tracking system are discussed and calculated, and the influence that the tracking system jitter inflicts on the target image contrast and SNR is analyzed in details. The dynamic optical transfer function of target tracking system with jitter is derived, and an image sharpen technology based on stray light transmission theory is proposed to remove blur caused by system jitter. Some practical methods used to suppress stray light and evaluate the performance of eliminating stray light are also presented in the dissertation.The low contrast target detection system is designed and implemented, and the performances of the improvement measures proposed in this dissertation are evaluated by real experiment. The model of stray light is analyzed and the improvement measures to eliminate stray light are put forward. Based on high-sensitivity image sensor technology, the detecting sensitivity, SNR characteristics and spectral response characteristics of some various cameras are discussed, and the camera that is suitable to detecting and tracking high speed target in low-contrast condition. Many image processing and target detecting algorithms, such as multi-image accumulation to restrain atmospheric effects, background reduction to improve the target imaging contrast, and spectral filtering to improve the effectiveness of the SNR, are also validated by target detection system. Background measurement instrument is developed to measure and calibrate the intensity of sky background radiation in certain space, and it can provide some crucial parameters for image background processing and low-contrast target detection by obtaining valuable sky background data. The image sharpness method based on incremental Wiener filtering with exponential law can enhance the imaging quality of the detections system in terms of contrast improvement at 20% and SNR improvement at more than 2.5dB.At last, the performances of low contrast target detecting and tracking system were summarized. Moreover, this dissertation prospects the future research focus, and the image processing algorithm should be studied by combination of detector performance and system dynamic transfer function, and photoelectric detection tracking mechanism and image sharpening techniques are explored using novel idea.