摘要
为了满足微小型载体负载能力的要求,利用BF548处理器体积小、功耗低以及外设资源丰富的优点,设计和验证了一种微小型低功耗图像融合实验平台.并针对所设计的硬件平台,提出快速实现的基于像素灰度概率密度分布离散度分析的红外与可见光图像融合算法.该算法利用灰度概率密度分布之间的离散度判断目标区域和背景区域,不同区域采用不同融合规则.仿真实验表明,该算法相对于传统提升小波融合算法,信息熵提高了16%,标准差提高了86%以上.
In order to meet the load capacity of the mini-system,the image fusion system should have low power dissipation and small volume. With the advantages of small volume,low power dissipation and plentiful peripherals,BF548 is selected as the main processor of the image fusion system to satisfy the special requirement of the mini-system. Then a new optimized fusion algorithm based K-L entropy analysis is proposed. The infrared source image is segmented into object and background region based on K-L entropy between visible and infrared images,and different fusion strategies are used for different regions. Through simulation experiment,it is proved that the information entropy based on this algorithm improves 16% over that based on traditional lifting wavelet algorithm,and the standard deviation improves more than 86%.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2010年第6期693-697,共5页
Transactions of Beijing Institute of Technology
基金
国家自然科学基金资助项目(60801050)
