信息密度与图像主观评价的关系研究

发布时间：2018-10-24 22:44

【摘要】：在遥感成像系统的研制过程当中,对其性能合理的预估有利于有效地指导系统的设计工作。调制传递函数(MTF)、信噪比(SNR)是较为常用的遥感成像系统的性能预估模型,但它们都无法全面反映遥感成像系统的综合性能。遥感成像系统的信息密度融合了MTF、SNR和边带混叠等多种像质表征参数,能够体现遥感成像系统多方面的性能。文章通过研究信息密度与图像主观评价的关系对信息密度用于预估遥感成像系统综合性能的合理性做了相关的研究。首先,设计不同的遥感成像系统,使其具有不同的信息密度值,以信息密度来体现不同系统的性能优劣;然后,利用不同系统对同一场景进行成像仿真,并对仿真输出的图像进行了主观量化评分;最后,利用相关性分析的手段研究了不同信息密度及其对应的主观评分,研究结果验证了信息密度作为遥感成像系统的一种综合性能预估模型的合理性。
[Abstract]:In the process of development of remote sensing imaging system, reasonable prediction of its performance is helpful to guide the design of the system effectively. Modulation transfer function (MTF),) signal-to-noise ratio (SNR) is a commonly used performance prediction model for remote sensing imaging systems, but none of them can fully reflect the comprehensive performance of remote sensing imaging systems. The information density of remote sensing imaging system is a combination of MTF,SNR, sideband aliasing and other image quality characterization parameters, which can reflect the performance of remote sensing imaging system in many aspects. By studying the relationship between information density and image subjective evaluation, the rationality of information density in predicting the comprehensive performance of remote sensing imaging system is studied in this paper. First of all, different remote sensing imaging systems are designed so that they have different information density values to reflect the performance of different systems. Finally, the different information density and its corresponding subjective score are studied by means of correlation analysis. The results verify the rationality of the information density as a comprehensive performance prediction model of remote sensing imaging system.
【作者单位】： 北京空间机电研究所;
【基金】：国家重大科技专项工程
【分类号】：TP751
，

本文编号：2292789