作者: Qingping Li , Junping Du , Fuzhao Song , Chao Wang , Honggang Liu
DOI: 10.1109/CCDC.2013.6561609
关键词: Region growing 、 Pattern recognition 、 Color image 、 Range segmentation 、 Binary image 、 Feature detection (computer vision) 、 Image fusion 、 Segmentation 、 Image processing 、 Morphological gradient 、 Image restoration 、 Artificial intelligence 、 Computer vision 、 Image gradient 、 Feature extraction 、 Pixel 、 Segmentation-based object categorization 、 Image segmentation 、 Image texture 、 Scale-space segmentation 、 Minimum spanning tree-based segmentation 、 Computer science 、 Digital image processing
摘要: In image fusion domain, generic pixel-based methods are sensitive to imperfections of source images, and it therefore has much influence on the quality results. Focusing this problem, a region-based multi-focus scheme is proposed based local spatial frequency (LSF) in paper. Firstly, calculate LSF for each pixel segmentation average introduced segment images. From segmented image, shared region representation obtained label The identification important features representation, (RSF), used guide process. experimental results show that works well fusion, outperforms comparison terms visual objective evaluations.