作者: Xia-qing XIE , Quan-wei BAI , Lei HOU , Xu WU
DOI: 10.1016/S1005-8885(13)60209-5
关键词:
摘要: Abstract Image retrieval is increasingly necessary for multi-media resources to improve user experiences and expand approach. This paper studies into recent process made in semantic-based image (SBIR), establishes a model (SBIRM) ‘Classic Reading’ Platform. The chooses three important features from illustrations of books including color, texture shape extract feature vector form 31-dimentional vector. Vector normalization performed eliminate the difference between different similarity measurement used compare two images. Finally, K-nearest neighbor (KNN) was train vectors decide which book given belongs to. experimental result shows that precision this can be more than 90%.