基于聚类算法的视频内容识别研究

陈双全

电脑与电信 ›› 2017, Vol. 1 ›› Issue (11) : 44-46.

电脑与电信 ›› 2017, Vol. 1 ›› Issue (11) : 44-46.
应用技术与研究

基于聚类算法的视频内容识别研究

  • 陈双全
作者信息 +

Research on Video Content Recognition Based on Clustering Algorithm

  • CHEN Shuang-quan
Author information +
文章历史 +

摘要

本文研究了视频网站内容监播、监控的规律,并且依据新的规律改进了现有的K-MEANS聚类算法。受互视频内容数据非结构化因素的影响,K-MEANS聚类算法进行聚类识别和计算效率上存在明显的缺陷。K-MEANS聚类改进算法的优势在考虑到了视频帧序列关系,并保持了可接受的检索速度。

Abstract

In this paper, we study the rules of video website content monitoring, and improve the existing K-MEANS clustering algorithm according to the new rule. The K-MEANS clustering algorithm has obvious defects in clustering recognition and computing efficiency due to the influence of unstructured factors of the video content data. The advantage of the improved K-MEANS clustering algorithm takes into account the sequence relations of video frames and maintains the acceptable retrieval speed.

关键词

数据挖掘 / 视频内容识别 / 聚类分析 / K-MEANS聚类算法

Key words

data mining / video content recognition / clustering analysis / K-MEANS clustering algorithm

引用本文

导出引用
陈双全. 基于聚类算法的视频内容识别研究[J]. 电脑与电信. 2017, 1(11): 44-46
CHEN Shuang-quan. Research on Video Content Recognition Based on Clustering Algorithm[J]. Computer & Telecommunication. 2017, 1(11): 44-46
中图分类号: TP391.41   

Accesses

Citation

Detail

段落导航
相关文章

/