基于内容的图像检索系统.docVIP

摘要 近年来，随着多媒体技术和计算机网络的飞速发展，全世界的数字图像的容量正以惊人的的速度增长。这些数字图像中包含了大量有用的信息。然而，由于这些图像是无序地分布在世界各地，图像中包含的信息无法被有效的访问和利用。一般的文字有哪些信誉好的足球投注网站引擎对于图像数据来说，几乎束手无策，这就要求有一种能够快速而且准确的查找访问图像的技术，也就是所谓的图像检索技术。 基于内容的图像检索(Content-Based Image Retrieval，CBIR)指的是查询条件本身就是一幅图像，它建立索引的方式是通过提取图像特征，然后通过计算比较这些特征和查询条件之间的距离，来决定两幅图像的相似程度。 本文主要对基于内容的图像检索系统进行设计，系统设计的重点在于图像的低层特征提取。本系统设计主要包括：图像低层特征（颜色、纹理等特征）的提取及描述，多特征融合技术和特征匹配技术等方面。本文首先论述了国内外基于内容的图像检索技术的研究现状、检索特点以及应用领域，从图像检索系统的基本原理入手，最后，本用MATLAB GUIGraphical User Interface)设计实现了一个基于的图像检索系统，并对系统的各个模块及其功能进行了介绍。仿真实验，对Corel图像库中的500幅图像进行了检索实验。实验结果表明，本系统计算简单、有效，能够取得较好的检索效果。 Abstract In recent years, with the development of multimedia technology and the rapid development of computer network, the worldwide digital image capacity is increasing at an alarming rate. These digital image contains a lot of useful information. However, these images are randomly distributed throughout the world, the image contains information cannot be accessed and used effectively. General text search engine for the image data seems almost incapable, this requires a fast and accurate image searching technology, also known as the image retrieval technology. Content based image retrieval ( Content-Based Image Retrieval, CBIR ) itself is a pair of images, it is indexed by the image feature extraction, then determine the degree of similarity of two images through the calculation and comparison of the features and the distance between the query conditions. This paper focuses on the CBIR system design,. This design is focused on low-level feature extraction.of image .The system design mainly includes: low level image features ( color, texture feature extraction and description ) extraction and description, multiple features fusion and feature matching technology and so on. This paper first discusses the domestic and foreign content based image retrieval technology research status, characteristics and application field of retrieval, and then from the image retrieval system basic principle, realization of imag