图象检索算子开放测试平台T-Brief设计与实现

Design and Implementation of a Test Bed for Retrieving Image with Extracted Feature

基于内容的图象检索是近年来多媒体技术领域发展的一个热点之一，大量基于特征的检索算法不断涌现。该文介绍一个对算法开放的抽取特征检索图象的算法测试平台，该平台可以即时集成现有的多种不同算法，并便于管理，同时它还提供了诸如综合检索，渐进检索等功能，可用于算法研究，性能比较等。

Content Based lmage Retrieval (CBIR) has become one of the hotspots in the domain of multimedia technology in recent years. A mass of retrieval algorithms based on feature extraction comes forth. In this paper we present an algorithm-open test bed for retrieving image with extracted features (T-Brief). It can integrate various algorithms at hand, give a good manage, and provide functions such as compound retrieval, step-up retrieval,etc. It can also be used for algorithm evaluation and performance comparison.