期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2017
卷号：17
期号：2
页码：94-97
出版社：International Journal of Computer Science and Network Security
摘要：Feature extraction and presentation is of the most critical issues in content-based image retrieval (CBIR). CBIR system commonly extracts retrieval results respecting to the similarities of the extracted feature of the given image and the candidate images. In these systems, primary image features such as color, texture, shape and location are autonomously extracted and stored as a feature vector in a database for image comparison. This paper presents a new approach for content-based image retrieval based on color and texture data. In this regard, color histogram and color moment are used as color feature further, PCA statistical method is applied to reduce dimensions.
关键词：Content-based image retrieval (CBIR); search and retrieval; texture extraction; color moment
