摘要
针对视频图像大数据的快速增长,如何从视频中快速检索出敏感信息急待解决。本文提出了一种基于镜头边界的相似系数关键帧提取算法以获取视频关键帧,并设计基于深度学习的分类模型进行分类。最后通过实验对比选取最好的分类模型。
For the rapid growth of large data video images, how to quickly retrieve the sensitive informa- tion from the video to be resolved, In this paper, a key frame algorithm based on shot boundary similarity coefficient is proposed to obtain the key frame of the video, and the classification model based on the deep learning is used to classify the video key frames. Finally, the best classification model is selected through experiments.
作者
李想
LI Xiang(Wuhan Research Institute of Posts and Telecommunications, Wuhan 430000, Chin)
出处
《电子设计工程》
2017年第21期137-140,共4页
Electronic Design Engineering
关键词
深度学习
视频检索
关键帧
deep learning
video retrieval
key frame
