基于帧间特征矩阵的同源视频检测
Homologous Video Detection Based on Inter-Frame Feature Matrix
DOI: 10.12677/SEA.2022.111015, PDF,   
作者: 张雪梅, 康宇哲：浙江理工大学，浙江 杭州；沈学文：浙江传媒学院，浙江 杭州
关键词: 帧间时空特征；三帧差分法；帧间特征序列；Inter-Frame Feature Matrix； Three-Frame Difference Method； Inter-Frame Feature Sequence
摘要: 近年来，随着数字媒体软件的火热，网络中的近重复相似视频呈爆炸似增长，因此快速准确地筛选出海量视频中的同源视频是当下研究的重点课题。针对该课题，本文采取了基于帧间特征矩阵的同源视频检测方案，首先在视频帧间时空关系矩阵的基础上确定视频相应类别，然后进一步通过视频帧间特征序列对比来确认所检测视频是否与该类下的其他视频存在重复片段，并定位重复片段在视频中的位置。当重复片段占比超过一定阈值，即可判定被检测视频为同源视频。实验表明该方法在CC_WEB数据集上平均准确率可达93.2%，由此证明了该方法在保护视频知识产权领域的可用性。
Abstract: In recent years, with the popularity of digital media software, the nearly repeated similar videos in the network are exploding. Therefore, it is a key topic of current research to quickly and accurately screen out the homologous videos in the massive videos. For the subject, this paper adopted homologous video detection scheme based on characteristic matrix between frames. First, the corresponding categories of the video are determined on the basis of the space-time relationship matrix between the video frames, and then through the characteristics of sequence comparison between video frames to confirm whether the detected video has duplicate segments with other videos under this category, and locate the position of duplicate segments in the video. When the proportion of repeated clips exceeds a certain threshold, the detected video can be judged as homologous video. Experimental results show that the MAP (Mean Average Precision) of this method can reach 93.2% on CC_WEB data set, which proves the applicability of this method in the field of video intellectual property protection.
文章引用：张雪梅, 康宇哲, 沈学文. 基于帧间特征矩阵的同源视频检测[J]. 软件工程与应用, 2022, 11(1): 130-138. https://doi.org/10.12677/SEA.2022.111015

参考文献

[1] 顾佳伟, 赵瑞玮, 姜育刚. 视频拷贝检测方法综述[J]. 计算机研究与发展, 2017, 54(6): 1238-1250.
[2] 12426版权监测中心. 2019年中国网络版权监测报告(摘要) [N]. 中国新闻出版广电报, 2020-05-14(007).
[3] 王娜. 我国网络视频产业的版权困局与破解[J]. 当代电影, 2017(10): 193-195.
[4] 李小琛. 数字视频同源帧内复制–粘贴篡改取证研究[D]: [硕士学位论文]. 福州: 福建师范大学, 2018.
[5] 胡瑞娟. 网络舆情中的同源视频检测[D]: [硕士学位论文]. 天津: 中国民航大学, 2014.
[6] 张乃光, 李珊珊, 薛子育. 基于深度学习的盗版视频分类[J]. 广播电视信息, 2019(S1): 84-87.
[7] Hsu, C.C., Hung, T.Y., Lin, C.W., et al. (2008) Video Forgery Detection Using Correlation of Noise Residue. Proceedings of the 2008 IEEE 10th Workshop on Multimedia Signal Processing, Cairns, 8-10 October 2008, 170-174.
[8] 黄添强, 吴铁浩, 袁秀娟, 陈智文. 利用模式噪声聚类分析的视频非同源篡改检测[J]. 计算机科学与探索, 2011, 5(10): 914-920.
[9] 范清宇. 音视频数据获取与同源性分析关键技术研究[D]: [硕士学位论文]. 成都: 电子科技大学, 2018.
[10] Jiang, Y.G. and Wang, J. (2016) Partial Copy Detection in Videos: A Benchmark and an Evaluation of Popular Methods. IEEE Transactions on Big Data, 2, 32-42. [Google Scholar] [CrossRef] 
[11] 栗志磊, 李俊, 施智平, 姜那, 张永康. 用于视频行为识别的高效二维时序建模网络[J/OL]. 计算机工程与应用, 2021.
[12] 詹克羽, 孙岳, 李颖. 一种多尺度三维卷积的视频超分辨率方法[J]. 西安电子科技大学学报, 2021, 48(5): 8-14.
[13] Lin, T.Y., Dollár, P., Girshick, R., et al. (2017) Feature Pyramid Networks for Object Detection. Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, 21-26 July 2017, 2117-2125. [Google Scholar] [CrossRef] 
[14] Lecun, Y., Bottou, L., Bengio, Y., et al. (1998) Gradient-Based Learning Applied to Document Recognition. Proceedings of the IEEE, 86, 2278-2324. [Google Scholar] [CrossRef] 
[15] He, K.M., Zhang, X.Y., Ren, S.Q., et al. (2016) Deep Residual Learning for Image Recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, 27-30 June 2016, 770-778. [Google Scholar] [CrossRef] 
[16] Hara, K., Kataoka, H. and Satoh, Y. (2018) Can Spatiotemporal 3D CNNS Retrace the History of 2D CNNS and ImageNet? Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 6546-6555. [Google Scholar] [CrossRef] 
[17] Kordopatis-Zilos, G., Papadopoulos, S., Patras, I., et al. (2019) ViSiL: Fine-Grained Spatio-Temporal Video Similarity Learning. Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision, Seoul, 27 October-2 November 2019, 6351-6360.
[18] 樊丽萍. Chamfer距离特征提取的数据匹配算法研究[D]: [硕士学位论文]. 南昌: 南昌大学, 2012.
[19] 孙挺, 齐迎春, 耿国华. 基于帧间差分和背景差分的运动目标检测算法[J]. 吉林大学学报(工学版), 2016, 46(4): 1325-1329.
[20] 王霏. 基于视频的运动目标检测算法研究[D]: [硕士学位论文]. 长春: 吉林大学, 2014.
[21] 亢洁, 李晓静. 基于均值背景与三帧差分的运动目标检测[J]. 陕西科技大学学报, 2018, 36(1): 148-153
[22] 王飞飞. 基于稀疏自动编码器的近重复视频检索[J]. 电子技术与软件工程, 2017(3): 194-196.
[23] Chen, H., Zheng, L., Al Kontar, R., et al. (2020) Gaussian Process Inference Using Mini-Batch Stochastic Gradient Descent: Convergence Guarantees and Empirical Benefits.
[24] Wu, X., Hauptmann, A.G. and Ngo, C. (2007) Practical Elimination of Near-Duplicates from Web Video Search. Proceedings of the 15th ACM international conference on Multimedia, Augsburg, 25-29 September 2007, 218-227. [Google Scholar] [CrossRef] 

为你推荐