提出了建立一个WWW图象搜索引擎的方案，搜索引擎在Internet上浏览主页，对遇到的图象进行特征计算，抽取出索引信息，存储索引向量，根据图象内容进行相似图象的查找。分析了颜色直方图、边缘方向直方图、纹理分析和形状不变性等基于图象内容查询的技术，并对WWW图象搜索引擎的发展作出了预测。

This papers proposes the development of a world wide web image search engine, which browses homepages on Internet, computes the features of images embedded in the pages, extracts the index information based on image content, stores the index vectors, and a nswers the users' queries of similar images. The paper also analyzes the technologies of query be image content such as color histograms, edge orientation histograms, texture measures and shape invariants. In the ending discussion the requirement of world wide web image search engine is predicted.