多特征融合的视频镜头分割  

Video Shot Segmentation Based on Multi-feature Fusion

作　　者：来毅 辛可嘉 刘颖 LAI Yi;XIN Kejia;LIU Ying(School of Telecommunication and Information Engineering,Xi′an University of Posts and Telecommunications,Xi′an 710121,China;Key Laboratory of Spectral Imaging Technology,Chinese Academy of Sciences,Xi′an 710119,China)

机构地区：[1]西安邮电大学通信与信息工程学院,西安710121 [2]中国科学院光谱成像技术重点实验室,西安710119

出　　处：《电讯技术》2018年第7期792-797,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(41504115);陕省教育厅专项科研计划项目(17JK0707);模式识别国家重点实验室开放课题(201700013);中国科学院光谱成像技术重点实验室开放基金课题(LSIT201709D);浙江大学CAD&CG国家重点实验室开放课题(A1820)

摘　　要：针对传统镜头分割算法特征单一、分割效果较差等问题,提出了一种多特征融合的视频镜头分割算法。首先,通过建立一种优化模型,融合空间差异度量和感知哈希度量两种重要特征,构造一种镜头边界区分能力更强的镜头边界检测特征,即像素差异度量;其次,结合直方图差异度量,形成有效的镜头边界检测策略,以便更加准确判断是否发生镜头切换。实验结果表明,与传统算法相比,所提算法取得了较好的分割效果,具有较高的查准率和查全率。In order to deal with the problems of the feature singularity and the poor performance of shot segmentation existing in traditional algorithms,this paper presents a shot segmentation method based on multi-feature fusion.Firstly,in order to build pixel difference metric which is a kind of shot boundary detection feature with more discriminative ability,the optimization model is established that integrates two important features including the spatial difference metric and the perceptual hashing metric.Secondly,the histogram difference metric is combined with the histogram difference metric,and then the effective strategy for shot boundary detection is established to more accurately determine whether the shots switch.The experimental results demonstrate that the proposed method achieves better performance with high precision and recall compared with the traditional algorithms.

关 键 词：镜头分割 特征融合 空间差异度量 感知哈希度量 直方图差异度量 

分 类 号：TN911.73[电子电信—通信与信息系统]