摘要
随着社会和科技的发展,人们对基于内容的图像检索(CBIR)的要求越来越高,但因研究起点比效晚,技术上还不是很成熟。文中就图像检索所用到颜色、纹理、形状、轮廓的基本颜色特征中比较重要的颜色特征为主要研究对象,从传统的基于颜色直方图及其后来发展的主要面积直方图和平均面积直方图等思路出发,分析其优缺点,提出一种基于图像分割主要面积直方图方法,提高了图像检索的性能。
With the social and technological development, the requirements of people's content- based image retrieval (CBIR) is higher and higher,but the starting point of research is later,it is not very mature in technology. In this paper, the image retrieval using color, texture, shape, color characteristics of the basic outline of the more important for the color of the main characteristics of subjects, from the traditional color histogram and later on the main area of histograms and the average area of ideas, such as histogram based analysis of their advantages and disadvantages, a major area of image- based division histogram ways to improve performance of image retrieval.
出处
《计算机技术与发展》
2009年第4期38-40,共3页
Computer Technology and Development
基金
安徽省自然科学基金项目(050420204)
安徽省高校拔尖人才基金项目(05025102)