摘要
针对新一代遥感图像辐射分辨率提高的情况,设计了计算双字节联合熵的算法,讨论了量化级数对信息熵计算的影响,并考虑信源的记忆性质,对多波段图像讨论削减像元之间相关性及多波段间相关性的方法,进而对接近无记忆信源的图像数据进行联合熵的计算和分析,取得更为合理的评价结果。所发展的方法适用于多波段影像融合中最适宜的波段选择和质量评估。
Aimming at the improvement of radiation resolution in new generation remote sensing images, a method that used for computing double byte joint entropy is provided. At the same time, some further questions, such as how the quantitative series affect the entropy, what kind of method can be used to reduce the correlation of pixels and different bands in multispectral image and how to compute and analyse joint entropy for images which is close to non-memory for more accurate result, are discussed. The developing method can be used for the optimum bands selecting and quality evaluation in image fusion.
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2006年第6期476-480,共5页
Geomatics and Information Science of Wuhan University
基金
国家自然科学基金资助项目(40171032)
