基于内容的图像检索技术的研究与实现

Study on content-based image retrieval technology and its implementation

如何为内容丰富多变的大量图像数据编制索引并利用该索引进行高效地相似检索是研究的核心问题。相似图像检索系统通过图像特征提取器提取图像的特征,提供访问图像内容的方法;距离函数是用来计算这些特征之间相似程度的主要工具。实验证明,该系统可以高效地为用户检索出指定特征的图像,对实际应用具有重要的价值。

This paper mainly investigates how to index the images in database and use the index to retrieval the image efficiently. The similarity image retrieval system designed in this paper extracts the image＇ s feature using the feature extractor. And the feature extractor provides a method to access the content of images. The distance function is the main tool to distinguish the similarity between the features. The experiment proves that the system can effectually give search results which have the given feature and it has important application value.