人工智能算法在基于内容的视频检索中的应用  

The application of artificial intelligence algorithms to content-based video indexing and retrieving system

作　　者：李雪梅[1] 王健敏[1] 丁峰[2] 

机构地区：[1]仲恺农业技术学院,广东广州510225 [2]华南理工大学,广东广州510640

出　　处：《机电工程技术》2005年第6期47-49,共3页Mechanical & Electrical Engineering Technology

摘　　要：随着计算机技术的发展,视频数据的容量以几何级数增长。如何在海量视频数据库中快速地检索到所需片断成为热门的研究课题。本文提出了将人工智能算法应用到基于内容的视频检索系统中,利用反馈式人工神经网络和自适应匹配算法,对视频进行基于内容的分类、索引及检索;通过对图像特征的提取与人工智能相结合的思想和方法,实现了图像快速有效的分类、索引和检索。最后对该方法进行了讨论。With the development of computer technology, the capacity of video data is increased dramatically during these days. How to efficiently index and retrieve the desired video segments from the vast video database becomes a pop research topic. A new way that applies mature artificial intelligence algorithms to content-based video indexing and retrieving system to enhance the system's performance is introduced in this paper. Using the back propagation neural network and self-organizing mapping algorithm, the video data are sorted, indexed and retrieved on the basis of the video's content. These artificial intelligence algorithms, combined with the texture extraction algorithms, perform well in the classifying, indexing and retrieving the video segments. At the end of the paper some discussions are presented.

关 键 词：基于内容的视频图像检索 特征提取 反馈式人工神经网 自适应匹配算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]