星 载 的 单 像 素 运 动 成 像 影 响 分 析 与 补 偿

Single-pixel imaging is a new type of imaging technology which uses a non-scanning single-pixel detector to image objects and has attracted much attention from the public since it was produced. It has forged ahead from theory research to experiment exploration and industrial appliance after nearly thirty years of development. What’s more, it has profound potential value in micro-medicine, remote sensing, statement detection, and other aspects. Single-pixel imaging projects objection via speckle light field of consecutive projection. Light intensity is detected by a single-pixel detector rather than a traditional area detector and image restoration requires correlation calculation between speckle light field and detection information. Single-pixel imaging can be divided into forward and backward modulation modes according to the different modulation modes. In the forward modulation mode, speckle light field is created by the spatial light modulator or light source array and used to illuminate the object, then the reflected light intensity is collected by a non-scanning single-pixel detector. In the backward modulation mode, the image of the object is sampled by the spatial light modulator, and the corresponding light intensity is detected by a non-scanning single-pixel detector. Forward modulation is more valuable in remote imaging detection, therefore, it is applied in this study. Since single-pixel imaging uses a non-scanning single-pixel detector to obtain spatially resolved information, it requires a large amount of different modulation information from spatial light modulator or light source array at different times. Therefore, single-pixel imaging sacrifices temporal resolution in exchange for spatial resolution. Generally, to restore high-quality object images with single-pixel imaging, a large number of speckle light patterns are needed to illuminate the object and the corresponding intensity are measured sequentially. This mode is effective in the static object imaging. However, when applying single-pixel imaging to handle a moving object directly, the images restored are likely to be disturbed by motion blur. This is mostly because the position of the moving object is changing while measuring the reflected light intensity, which is different from the case in static object. The traditional imaging system can make moving object in a relatively static state and record a clear object image by shortening the single exposure time. The existing high-speed camera can shorten the single exposure time to less than 1/10 000 seconds. It can photograph the crack propagation at the moment of glass breakage or a bullet in flight. However, single-pixel imaging is different from traditional imaging. The single-frame imaging time of it is difficult to shorten to this level while requiring high imaging quality. If the single-pixel imaging is used in satellite detection, the correlation between reference light field and detector signal will be lost due to the orientation dither caused by the change of satellite attitude, which seriously degrades the reconstructed image quality. Therefor we propose a tracking compensation scheme which can extend single-pixel imaging from continuous exposure sampling imaging of static objection to single-pixel imaging of moving objection. There are two studies in this paper, impact analysis of high quality spaceborne single-pixel motion imaging reconstruction and compensation of spaceborne single-pixel motion imaging based on it. First of all, to solve the above two problems, single-pixel imaging system and its theory are introduced in detail, which lays a theoretical and experimental foundation for the further introduction of solutions and verification. The known series of structured speckles and the measured light intensity can be combined and inverted using a variety of algorithms to yield a good estimated image of the object. Differential ghost imaging is used to complete our research because of its good denoising ability. Second, a tracking compensation scheme in light field was proposed to solve the influence of the satellite orientation dither on the imaging. By analyzing the influence of the roll, pitch and yaw angles on correction imaging, it is clear that the reason of image quality degradation is lower connectively between the speckle light field and the detection value of single-pixel detector due to the satellite orientation dither. The correlation can be restored by light field tracking compensation scheme, and the simulation results show that the scheme is suitable for roll, pitch and yaw angles motion. The experimental results show that the peak signal-to-noise ratio of the reconstructed image is improved obviously, which proves the effectiveness of the method. Finally, the light field tracking compensation scheme of satellite orientation dither is carried out for typical scenes. Comparing the results before and after compensation, the PSNR is increased by 4.7 dB at least, which effectively improves the imaging quality. This scheme provides an effective technical approach for single-pixel imaging of space-borne motion.