Abstract: Content-based image/video reranking is a new technique to refine the original text-based search result by incorporating image’s visual information.It aims at improving the search performance and enhancing the search experience of users through some certain methods.It is a new way for Internet multimedia image/video search.This paper analyzes the status of reranking technique systematically，discusses the characteristics of various types of reranking technique and applications in detail，and summarizes the existing evaluation methods and databases，and some trends in this field are also presented in the end.

Key words: content-based reranking, image/video search, relevance reranking, diversity reranking

摘要： 基于内容的图像/视频搜索重排序技术是指在基于文本的图像/视频搜索结果的基础上，利用数据的视觉信息，通过某种方法对原始搜索结果重新排序的过程，目的是提高搜索质量和提升用户搜索体验，是一种互联网多媒体图像/视频搜索新模式。对这一技术进行了综述，系统地分析了重排序技术的发展现状，详尽地探讨了各类重排序技术的特点及应用，总结了现有评价方法和数据库，指出了当前重排序技术的发展趋势。

关键词: 基于内容的重排序, 图像/视频搜索, 相关性重排序, 多样性重排序