论文部分内容阅读
卫星搭载的有效载荷在进行在轨对地观测过程中,受仪器本身特性和地球固有特性因素的影响,在图像上会产生多种观测失准和干扰现象。如图像旋转现象,蝴蝶结效应以及探元非均匀性条纹现象等。针对这类可以预见到的冗余和条纹,对仪器设计优化以及借助卫星星历轨道等辅助数据和成熟的图像处理算法可以对其实现有效地恢复。仪器在轨运行期间新引入的周期性或非周期性干扰,则需综合运用多种数据来源和关于干扰产生原理的先验知识,并结合图像处理手段实现图像恢复,同时保存原有辐射信息的真实性,为定量研究提供高质量可靠的数据准备。本研究针对某卫星中波红外图像干扰条纹,利用星上定标数据,并通过条纹现象分析和产生机理分析,探索了一种新的基于空域信号补偿原理的图像干扰条纹处理算法,通过大量图像数据处理实践证明,该算法对图像质量改善明显,且在动态范围和辐射信息保存方面优于传统频域滤波算法。
In carrying out the on-orbit-observation process, the payload carried by satellites is affected by the characteristics of the instrument itself and the inherent characteristics of the earth, resulting in a variety of observation misalignments and interferences on the images. Such as image rotation phenomenon, bow effect and probe non-uniformity streak phenomenon. For such predictable redundancy and streaks, instrument design optimization as well as auxiliary data and sophisticated image processing algorithms such as satellite ephemeris can be effectively restored. The periodic or non-periodic interference newly introduced during the orbital operation of the instrument requires the integrated use of multiple data sources and prior knowledge about the principle of interference generation, combined with image processing to achieve image restoration while preserving the original radiation information Authenticity, providing high quality and reliable data preparation for quantitative research. In this paper, a new method of image interference fringe processing based on space-borne signal compensation is proposed, which is based on the calibration data of satellite and the streak phenomenon analysis and mechanism analysis. Data processing practice proves that this algorithm has obvious improvement on image quality and is better than the traditional frequency domain filtering algorithm in terms of dynamic range and radiation information preservation.