卫星在轨MTF测评研究及应用

【作者】 王鸿南；

【导师】 夏德深；

【作者基本信息】 南京理工大学 ， 计算机应用技术， 2004， 硕士

【摘要】 从卫星传回的遥感图像上获取卫星遥感器的关键评价指标MTF的变化情况，是目前国际上一个非常活跃的研究课题，对于在轨运行的卫星遥感器的监测和寿命预估，以及后继卫星的研制，有着非常重要的意义。目前，国内在这方面的研究尚处在起步阶段。因此，在本文中将着重介绍如何从卫星遥感图像获取在轨卫星MTF的基本方法及在MTF测评研究中的一些创新应用，其中如何从卫星遥感图像获取在轨卫星MTF的基本方法主要是在借鉴美国NASA(航空航天局)方法的基础上，总结并设计了适合本项目研究的方法，并在文中给出以中巴地球资源卫星遥感图像做的实验结果。本项研究中的一些创新应用则是在研究MTF理论及实验中发现和改进的两种新方法，可以分别应用于图像清晰度评价和图像恢复，其中图像清晰度的评价方法已被用于卫星图像质量评价的软件中，并且该方法将在图象图形学报上发表。 本文首先介绍和理解MTF的基本原理及从卫星遥感图像获取在轨卫星MTF的基本方法(刀刃法和脉冲法)，继而选择合适的图像进行试验和分析。其次，介绍图像清晰度评价的新方法——点锐度方法和基于MTF光学理论的图像恢复。更多还原

【Abstract】 To obtain the MTF which is the key indicator of the digital imaging system’s performance, is a active researching topic in the satellite imaging system area. Measurement of the MTF for digital imaging systems is problematic because of sampling and consequent aliasing issue. But the result of the research is very important to supervise the on-obit satellite and develop the next generation satellite. Nowadays, domestic research of this area is just at beginning, so there are still many problems and difficulties in research work. Thus, the work of this paper are mainly focus on how to obtain the MTF form on-orbit satellite and two new method proposed based on the MTF theory. The way to obtain MTF is mainly derive from the method which NASA had adopted, and we improved it to make this method became more compatible to our satellite. When it comes to the two new methods, one is apply to the digital image definition assessment and the other is apply to reconstruction for digital image. Especially, the former assessment method called Point acutance arithmetic is already adopted by the authorities and will be published in the Journal of Image and Graphic.In this paper, firstly, we will introduce both the base theory of MTF and the method to obtain the MTF form on-orbit satellite. Secondly, we will select some image to experiment and give the analytic result. Finally, we will give the two new methods which respectively apply to image definition assessment and image reconstruction.更多还原