基于CEDD特征的图像搜索引擎设计与实现
首发时间:2014-12-29
摘要:基于内容的图像检索是计算机视觉技术在图像检索方面的应用,这种检索方式主要面向图像内容,通过对图像的颜色、形状、纹理等信息进行分析,建立图像索引。用户上传搜索图片后,系统将被检索图像与图像库进行特征比对,返回相似图像。本文设计的图像搜索引擎系统是对基于内容图像检索系统的扩展,添加了定制的图像爬虫模块。CEDD(Color and Edge Directivity Descriptor)全称为颜色和边缘方向性描述符。该描述符结合了图像颜色和纹理两方面的特征。本文在CEDD特征进行深入分析和研究的基础上,使用CEDD特征进行图像索引的建立,不仅降低了索引建立的时间,而且提高了图像搜索效率。?
For information in English, please click here
Design and implementation of image search engine based on CEDD
Abstract:Content-based image retrieval is an application of computer vision technology in terms of image searching. This retrieval technology is mainly for image content, through the analysis of its features, such as color, shape, texture and other information of image, an appropriate image index will be established. After the searching picture is uploaded and the feature match procedure is executed, similar images result will be returned to users. The image search engine designed by this paper is an expansion of Content-based image retrieval system, with a custom picture crawling module in addition. This paper conducts a deep research on CEDD which incorporates color and texture information in a histogram. CEDD is short for "Color and Edge Directivity Descriptor. By using CEDD, the image search engine needs less time to make index on image and acquires a higher search accuracy compared with the previous.
Keywords: Search Engine Image Retrieval CEDD Image Indexing
基金:
论文图表:
引用
No.4622622102431514****
同行评议
共计0人参与
勘误表
基于CEDD特征的图像搜索引擎设计与实现
评论
全部评论0/1000