基于ViBe算法运动特征的关键帧提取算法

doi:10.6040/j.issn.1672-3961.0.2019.276

摘要/Abstract

摘要：

针对视频关键帧提取算法中运动类视频运动目标特征不易提取所造成的错选和漏选问题,提出一种基于背景建模(visual background extractor, ViBe)算法的前景运动目标特征提取的关键帧提取算法。通过ViBe算法对视频序列进行前景目标检测,提取前景运动目标的尺度不变特征变换(scale invariant feature transform, SIFT)特征,并对相邻帧之间的特征数据进行特征点匹配,根据定义的公式计算视频帧的相似度,然后根据提出的关键帧判别方法输出视频的关键帧。试验结果表明,该算法能较好的解决运动类视频关键帧提取中出现的漏选和错选问题,与基于SIFT分布直方图的算法相比,其查准率和查全率的综合指标F₁值有较好提高。因此该算法对于判别运动类视频中包含关键动作的关键帧具有较好的检测效果。

关键词: 关键帧提取, 视频序列背景检测算法, 尺度不变特征变换, 目标特征, 特征点匹配

Abstract:

Aiming at the fact that the background was dominant in the key frame extraction algorithm, in which the foreground target was too small and it was not easy to extract the features of moving targets in sports video, a key frame extraction algorithm for foreground moving target feature extraction based on background modeling algorithm was proposed, which was called visual background extractor (ViBe) algoritm. The foreground target detection of video sequence was firstly carried out using ViBe algorithm, afterwards the scale-invariant feature transformation (SIFT) features of the foreground moving target were extracted. Based on the similarity calculated from video frame series, the key frames of video were output according to the key frame discrimination method. The experimental results showed that the proposed algorithm could solve the problem of missed selection and misselection in traditional key frame extraction. Compared with the algorithm based on SIFT distribution histogram, the F₁ score was well improved. The algorithm based on ViBe could effectively identify key frames in sports video.

Key words: key frame extraction, background modeling, scale invariant feature transform, target characteristic, feature point matching

中图分类号:

TP37

李秋玲,邵宝民,赵磊,王振,姜雪. 基于ViBe算法运动特征的关键帧提取算法[J]. 山东大学学报 (工学版), 2020, 50(1): 8-13.

Qiuling LI,Baomin SHAO,Lei ZHAO,Zhen WANG,Xue JIANG. Key frame extraction based on ViBe algorithm for motion feature extraction[J]. Journal of Shandong University(Engineering Science), 2020, 50(1): 8-13.

图/表 12

图1

图2

图3

图4

表1

图5

图6

图7

图8

表2

表3

图9

参考文献 20

1	张佳.体育视频切变检测与关键帧提取[D].湖北:华中科技大学, 2014.
	ZHANG Jia. Sports video cut detection and key frame extraction[D]. Hubei: Huazhong University of science and technology, 2014.
2	ZONG Z, GONG Q. Key frame extraction based on dynamic color histogram and fast wavelet histogram[C]//2017 IEEE International Conference on Information and Automation (ICIA). Macau, China: IEEE, 2017: 183-188.
3	郝会芬.视频镜头分割和关键帧提取关键技术研究[D].湖北:华中科技大学, 2015.
	HAO Huifen. Key technology research of video shot segmentation and key frame extraction[D]. Hubei: Huazhong University of Science and Technology, 2015.
4	罗森林, 马舒洁, 梁静, 等. 基于子镜头聚类方法的关键帧提取技术[J]. 北京理工大学学报, 2011, 31 (7): 351- 352.
	LUO Senling , MA Shujie , LIANG Jing , et al. Method of key frame extraction based on subshot clustering[J]. Transactions of Beijing Institute of Techology, 2011, 31 (7): 351- 352.
5	将元友. 一种基于聚类的关键帧提取算法[J]. 数字技术与应用, 2014, (11): 126- 127.
	JIANG Yuanyou . A key frame extraction algorithm based on clustering[J]. Digital Technology and Application, 2014, (11): 126- 127.
6	胡圆圆.基于视觉显著性的视频关键帧提取与帧速率上转换[D].南京:南京邮电大学, 2016.
	HU Yuanyuan. Video keyframe extraction and frame rate up conversion based on vision saliency[D]. Nanjing: Nanjing University of Post and Telecommunication, 2016.
7	SHI Lichun , CAI Jingzhi , ZHANG Mingxin . Key frame extraction algorithm based on rough set in compressed domain[J]. Computer Engineering, 2011, 33 (10): 2340- 2346.
8	白慧茹, 吕进来. 基于聚类方法改进的关键帧提取算法[J]. 计算机工程与设计, 2017, 38 (7): 1929- 1933.
	BAI Huiru , L Jinlai . Improved algorithm of key frame extraction based on clustering methods[J]. Computer Engineering and Design, 2017, 38 (7): 1929- 1933.
9	沈济南, 梁芳, 郑明辉. 基于自主扰动变异差分视频的关键帧提取算法[J]. 武汉大学学报, 2014, 60 (5): 434- 440.
	SHEN Jinan , LIANG Fang , ZHENG Minghui . Improved keyframe extraction algorithm based on self perturbation mutation differential evolution[J]. Journal of Wuhan University, 2014, 60 (5): 434- 440.
10	WOLF W. Key frame selection by motion analysis [C]//Proceedings of 1996 IEEE International Conferenceon Acoustics, Speech, and Signal Processing Conference Proceedings. New York, USA: IEEE, 1996: 1228-1231.
11	BARNICH O , DROOGENBROECK M V . ViBe: A universal background subtraction algorithm for video sequences[J]. IEEE Transactionson Image Processing, 2011, 20 (6): 1709- 1724. doi: 10.1109/TIP.2010.2101613
12	田丽华, 张咪, 李晨. 基于运动目标特征的关键帧提取算法[J]. 计算机应用研究, 2019, 36 (10): 3183- 3186.
	TIAN Lihua , ZHANG Mi , LI Chen . Key frame extrac-tion algorithm based on feature of moving target[J]. Application Research of Computers, 2019, 36 (10): 3183- 3186.
13	张文雅, 徐华中, 罗杰. 基于ViBe的复杂背景下的运动目标检测[J]. 计算机科学, 2017, 44 (9): 304- 307.
	ZHANG Wenya , XU Huazhong , LUO Jie . Moving objects detection under complex background based on ViBe[J]. Computer Science, 2017, 44 (9): 304- 307.
14	VAN M DROOGENBROECK and PAQUOT O. Background subtraction: experiments and improvements for vibe[C]//Proceedings of 25th IEEE International Conferenceon Computer Vision and Pattern Recognition. Workshops, RI, USA: IEEE Computer society Press, 2012: 32-37.
15	BARBIERI T, GOULARTE R. KS-SIFT: a keyframe extraction method based on local features[C]//Proceedings of IEEE International Symposium on Multimedia. Sao Paulo, Brazil: IEEE, 2015: 13-17.
16	李海洋, 文永革, 何红洲. 一种改进的SIFT特征点检测方法[J]. 计算机应用与软件, 2013, 30 (9): 147- 150. doi: 10.3969/j.issn.1000-386x.2013.09.041
	LI Haiyang , WEN Yongge , HE Hongzhou . An imporved SIFT feature point detection method[J]. Computer Applications and Software, 2013, 30 (9): 147- 150. doi: 10.3969/j.issn.1000-386x.2013.09.041
17	HU Xuelong, TANG Yingcheng, ZHANG Zhenghua. Video object matching based on sift algorithm[C]//Conference Neural on Networks and Signal Processing. Nanjing, China: IEEE, 2008: 412-415.
18	屈有佳.基于SIFT特征的关键帧提取算法研究[D].北京:北京交通大学, 2015.
	QU Youjia. Study of keyframe extraction algorithm based on SIFT fratures[D]. Beijing: Beijing Jiaotong University, 2015.
19	柳雪.视频检索中基于多特征的关键帧提取算法研究[D].江苏:中国矿业大学, 2015.
	LIU Xue. Research on keyframe extraction algorithm based on Multi-featurein video retrieval[D]. Jiangsu: China University of Mining and Technology, 2015.
20	HANNANE R , ELBOUSHAKI A , AFDELK , et al. An efficient method for video shot boundary detection and key frame extraction using SIFT-point distribution histogram[J]. International Journal of Multimedia Information Retrieval, 2016, 5 (2): 89- 104. doi: 10.1007/s13735-016-0095-6

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed

视频名称	总帧数/帧	镜头数/个	时长/s
体操视频	7 594	55	303.76
足球视频	5 124	17	204.96
羽毛球视频	3 750	19	50.00
篮球视频	4 432	19	177.28
击剑视频	2 762	24	110.48

[1]	逯跃锋,张奎,刘硕,吴跃,赵硕,李强,冯晨. 一种基于斜率差和方位角的矢量数据匹配算法[J]. 山东大学学报(工学版), 2016, 46(6): 31-39.
[2]	牟春倩,唐雁. 融合整体和局部信息的三维模型检索方法[J]. 山东大学学报(工学版), 2016, 46(6): 48-53.
[3]	张训华1,业宁2,王厚立3. 基于Harris角点的木材CT图像配准[J]. 山东大学学报(工学版), 2010, 40(5): 101-104.