首页 | 本学科首页   官方微博 | 高级检索  
     

基于K均值聚类的视频关键帧提取技术研究
引用本文:张一凡,李家辰,旷远有,刘盼,赵子元.基于K均值聚类的视频关键帧提取技术研究[J].电脑与信息技术,2021,29(1):13-16.
作者姓名:张一凡  李家辰  旷远有  刘盼  赵子元
作者单位:北方工业大学信息学院,北京 100144;北方工业大学信息学院,北京 100144;北方工业大学信息学院,北京 100144;北方工业大学信息学院,北京 100144;北方工业大学信息学院,北京 100144
基金项目:本项目受北京市大学生科学研究;创业行动计划资助
摘    要:文章提出了一种基于k均值聚类的视频关键帧提取算法.该算法在视频的镜头分割算法之上,通过层次聚类对视频内容进行初步划分,之后结合k均值聚类算法对初步提取的关键帧进行直方图特征对比去掉冗余帧,最终确定视频的关键帧序列.经过大量的实验数据证明,该算法能够提取出冗余度较低、代表性较强的关键帧,一定程度上为视频关键帧提取提供了有...

关 键 词:视频镜头分割  关键帧提取  k均值聚类

Research on video Key Frame Extraction Based on K-means Clustering
ZHANG Yi-fan,LI Jia-chen,KUANG Yuan-you,LIU Pan,ZHAO Zi-yuan.Research on video Key Frame Extraction Based on K-means Clustering[J].Computer and Information Technology,2021,29(1):13-16.
Authors:ZHANG Yi-fan  LI Jia-chen  KUANG Yuan-you  LIU Pan  ZHAO Zi-yuan
Affiliation:(School of Information Science and Technology,North China University of Technology,Beijing 100144,China)
Abstract:Based on the k-means clustering method,this paper proposes an algorithm for video key frame extraction.The algorithm,including video shots segmentation algorithm,partitions the video content through hierarchical clustering,then combined with k-means clustering algorithm for preliminary extract key frames histogram feature comparison to remove redundant frames,ultimately determine the key frames of video sequences.After a lot of testing,it is proved that the key frame redundancy extracted by this algorithm is relatively small,with high representativeness,and has certain validity of video key frame extraction.
Keywords:video shots segmential  key frame extaction  k-means clustering
本文献已被 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号