基于分布式的直方图检索方法研究及实现

Design and Application of Distributed Image Retrieval Based on Color Model to Histogram Conversion

基于内容的图像检索在检索效率和检索性能等方面一直存在着限制与不足。为提高图像检索效率,对基于内容的图像检索和分布式计算进行研究,提出一种基于图像颜色模型向直方图转换的分布式检索方法DHCIR(Distributed Image Retrieval method based on Color Model to Histogram Conversion),并基于该方法进行系统设计及实现。通过实际测试,该算法能够提供稳定、快速、高效的图像检索服务,提高图像检索的计算效率与准确性。

The retrieval efficiency and the retrieval capability are considered as two major problems existing in CBIR systems. In order to improve the retrieval efficiency,the authors present a new CBIR application based on large databases in distributed environment using DHCIR （Distributed Image Retrieval method based on Color Model to Histogram Conversion） algorithm by reviewing past researches in CBIR and distributed computing. A new image retrieval system using this method is also developed. The test results show this application can provide stable, fast and efficient image retrieval functions to improve the retrieval speed and accuracy.