基于QPSO和统计特征的红外和可见光图像融合.pdfVIP

第36 卷 第11 期 红 外 技 术 Vol.36 No.11 2014 年11 月 Infrared Technology Nov. 2014 基于 QPSO 和统计特征的红外与可见光图像融合 孙新德，刘国梅，薄树奎 （郑州航空工业管理学院计算机科学与应用系 河南 郑州 450015 ） 摘要：为解决红外图像与可见光图像融合中融合图像质量不高问题，提出了一种结合量子粒子群优 化和邻域统计的图像融合方法。红外和可见光源图像经过多尺度分解成为低频和高频子带。对低频 子带，采用系数加权平均的融合策略，并通过量子粒子群优化方法搜索最优的融合权值；对于高频 子带，采用受邻域统计信息调制的系数比较取大融合策略,通过逆变换重构图像得到融合结果。实验 结果表明该算法能够很好地将红外图像与可见光图像进行融合，且融合效果优于其它一些算法。 关键词：图像融合；QPSO；邻域统计；融合策略 中图分类号：TP391 文献标识码：A 文章编号：1001-8891(2014)11-0900-05 Fusion Algorithm for Infrared and Visible Light Images Based on QPSO and Neighbor Statistic Features SUN Xin-de，LIU Guo-mei ，BO Shu-kui (Department of Computer Science and Application, Zhengzhou Institute of Aeronautical Industry Management, Zhengzhou 450015, China) Abstract ：Being aimed at the fusion quality problem of infrared and visible light images with the same scene, a novel fusion algorithm based on Quantum-behaved Particle Swarm Optimization (QPSO) and neighbor statistic features is proposed in this paper. The source images were decomposed with various scales and direction，then many subband coefficients were obtained. For the low frequency subband, the coefficients weighted average fusion strategy was applied, and the optimal weight values were obtained by QPSO. For the high frequency subband, the fusion strategy of coefficients comparison by neighbor statistic feature modulation was proposed. The fusion image was obtained by inverse transform. The experimental results show that the proposed algorithm can fuse infrared and visible images well and acqu