摘要
图像数据库容量的增长,需要研究高效的索引技术来支持快速相似性检索的要求。总结了图像数据库检索技术的发展轨迹和特点,针对基于内容的图像检索技术中的局限性,从计算机底层硬件的角度提出了基于内容检索的流水索引法。该方法将基于内容的图像检索技术与CPU流水线结构紧密结合,对检索算法进行优化,通过举例比较,说明可提高图像数据库基于内容检索的速度。
As the volume of image database grows, it is urgent to work on highly effective index technique to support fast similarity search in a very large database. By summarizing the development and characteristics of retrieval method for image database, aiming at the limitation of content-based image retrieval, a pipelining index method based on content retrieval is proposed from the perspective of computer low-level hardware. In this method the content-based image retrieval and the pipeline architecture of CPU are perfectly combined, and the retrieval algorithm is optimized. It is shown that this method can accelerate search based on content in image database by comparison and examples.
出处
《电子工程师》
2008年第11期42-45,共4页
Electronic Engineer
关键词
图像数据库
流水索引法
基于内容检索
image database
pipelining index method
content-based retrieval
