多源视频序列噪声抑制融合算法  

Multi-source video sequence fusion algorithm restraining noise

作　　者：王昕[1,2] 魏有利 李高略[2] 刘富[1] 

机构地区：[1]吉林大学通信工程学院,吉林长春130022 [2]长春工业大学计算机科学与工程学院,吉林长春130012

出　　处：《中南大学学报（自然科学版）》2013年第S2期391-395,共5页Journal of Central South University:Science and Technology

基　　金：吉林省科技厅自然科学基金资助项目(201215127)

摘　　要：提出一种新的红外与可见光视频序列噪声抑制融合算法。首先采用图像分块结合局部加权信息熵的方法对红外图像进行目标区域分割;然后对已配准的红外与可见光源图像利用本文作者提出的新融合规则仅在目标区域进行融合;最后将融合后的目标区域与可见光图像的背景相结合。与传统融合算法的对比实验表明:该算法在抑制噪声的同时能更好地实现两类源图像的优势互补,具有实时性高和对噪声的鲁棒性好等特点。A new fusion method of infrared and visible video fusion was proposed to suppress noise.Firstly the approximate target regions of each single-frame infrared image were detected by image blocking together with the local weighted information entropy.Secondly,the new fusion algorithm was proposed to fuse the registered infrared and visible image in the target region.Lastly the fused target regions were fused with the background regions of visible sequences.The comparison experiments with the traditional fusion algorithms show that the algorithm can realize the complementary advantages of the two types of source images with good noise robustness and real-time performance.

关 键 词：视频融合 方差加权信息熵 提升静态小波变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]