[1]刘嘉宇,陈 平.基于关键帧定位和时空图卷积的异常行为识别[J].机械与电子,2022,(01):48-53.
　LIU Jiayu,CHEN Ping.Abnormal Behavior Recognition Based on Key Frame Location and Spatial-temporal Graph Convolution[J].Machinery & Electronics,2022,(01):48-53.
点击复制

基于关键帧定位和时空图卷积的异常行为识别()
分享到：

机械与电子[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2022年01期
页码:
48-53
栏目:
自动控制与检测
出版日期:
2022-01-20

文章信息/Info

Title:
Abnormal Behavior Recognition Based on Key Frame Location and Spatial-temporal Graph Convolution
文章编号:
1001-2257 ( 2022 ) 01-0048-06
作者:
刘嘉宇; 陈 平
中北大学信息探测与处理山西省重点实验室,山西 太原 030051
Author(s):
LIU Jiayu ;  CHEN Ping
( Shanxi Key Laboratory of Signal Capturing and Processing , North University of China , Taiyuan 030051 , China )
关键词:
异常行为; 骨架检测; 关键帧; 骨架时空图; 瓶颈残差; 时空图卷积
Keywords:
abnormal behavior ;  skeleton detection ;  key frame ;  skeletal space-time diagram ;  bottleneck residual ;  spatial-temporal graph convolution
分类号:
TP391.4
文献标志码:
A
摘要:
为提高监控视频中行人异常行为检测效率,提出了结合关键帧定位和时空图卷积的异常行为识别方法。该方法在人体骨架关键点检测的基础上,采用关键点运动特性定位视频中行人异常行为关键序列,利用时空图卷积网络可以提取行人时空特征的优点,在关键帧序列上构建人体骨架时空图,同时建立基于瓶颈残差模块的时空图卷积网络行为识别模型,实现对监控视频中行人异常行为的高效识别。采用自建数据集和公开数据集对该方法有效性进行检验,结果表明,该键帧定位算法可高效实现异常行为定位,结合基于瓶颈残差模块时空图卷积网络,在减少时空图卷积网络计算复杂度的同时提升了网络性能,能够有效判断行人异常行为。
Abstract:
In order to improve the detection efficiency of pedestrian abnormal behavior in surveillance video , an abnormal behavior recognition method combining key frame positioning and spatio-temporal map convolution is proposed.Based on the detection of key points of human skeleton , this method uses key point motion characteristics to locate pedestrian anomalies in video behavior key sequence , using spatio temporal map convolutional network to extract the advantages of pedestrian spatiotemporal features , construct the human skeleton spatiotemporal map on the key frame sequence , and establish a spatio-temporal map convolutional network behavior recognition model based on the bottleneck residual module to realize monitoring video efficient identification of abnormal behaviors of pedestrians in China.Self-built data sets and public data sets are used to test the effectiveness of the method.The results show that the key frame positioning algorithm in this paper can efficiently locate abnormal behaviors , combined with the spatio temporal graph convolution network based on the bottleneck residual module , while reducing the? computational complexity of the spatio-temporal graph convolutional network , it improves the network performance and can effectively judge the abnormal behavior of pedestrians.

参考文献/References:


[ 1 ] 王曲,赵炜琪,罗海勇,等 . 人群行为分析研究综述[ J ] . 计算机辅助设计与图形学学报,2018 , 30 ( 12 ): 2353-2365.
[ 2 ] TU Z G , XIE W , QIN Q Q , et al.Multi-stream CNN : learning representations based on human-related regions for action recognition [ J ] .Pattern recognition , 2018 , 79 : 32-43.
[ 3 ]张怡佳,茅耀斌 . 基于双流卷积神经网络的改进人体行为识别算法[ J ] . 计算机测量与控制,2018 , 26 ( 8 ):266-269 , 274.
[ 4 ] LI Q , YANG W Z , CHEN X Y , et al.Temporal segment connection network for action recognition [ J ] . IEEE Access , 2020 , 8 : 179118-179127.
[ 5 ] WHANGBO T K , EUN S J , JUNG E Y , et al.Personalized urination activity recognition based on a recurrent neural network using smart band [ J ] .International neurourology journal , 2018 , 22( Suppl 2 ): 91-100.
[ 6 ] 窦雪婷,王硕,季鑫盛 . 基于改进 DNN LSTM 算法的车辆前方行人行为识别方法[ J ] . 计算机测量与控制,2019 , 27 ( 11 ): 175-179.
[ 7 ] DUA N , SINGH S N , SEMWAL V B.Multi-input CNN? GRU based human activity recognition using wearable sensors [ J ] .Computing , 2021 , 103 ( 3 ): 1461-1478.
[ 8 ] LI C L , CUI Z , ZHENG W M , et al.Spatio-temporal graph convolution for skeleton based action recognition [ C ]// Pro- ceedings of the 32nd AAAI Conference on Artificial Intelligence.USA : Sheila McIlraith , 2018 : 7444-7452.
[ 9 ] LIU C F , LI X L , LI Q , et al.Robot recognizing humans intention and interacting with humans based on a multi task model combining ST-GCN-LSTM model and YOLO model [ J ] .Neurocomputing , 2021 , 430 : 174-184.
[ 10 ] CAO X , LIU X Y , AN M Y , et al.Vision-based fall detection using dense block with multi-channel convolutional fusion strategy [ J ] .IEEE Access , 2021 , 9 : 18318-18325.
[ 11 ] TSAI M F , CHEN C H.Spatial temporal variation graph convolutional networks( STV GCN ) for skeleton-based emotional action recognition [ J ] .IEEE Access , 2021 , 9 : 13870-13877.
[ 12 ] 汪成峰,王庆,梅树立,等 . 基于插值小波关键帧提取的动作评价算法[ J ] . 计算机工程, 2017 , 43 ( 1 ): 309-315.
[ 13 ] 郑恩,林靖宇 . 基于图像质量约束的无序图像关键帧提取[ J ] . 计算机工程, 2017 , 43 ( 11 ): 210-215.
[ 14 ] WIDIARTO W , HARIADI M , YUNIARNO E M. Keyframe selection of frame similarity to generate scene segmentation based on point operation [ J ] .International journal of electrical and computer engineering , 2018 , 8 ( 5 ): 2839-2846.
[ 15 ] FU H X , SONG G Q , WANG Y C.Improved YOLOv4 marine target detection combined with cbam [ J ] .Symmetry,2021 , 13 ( 4 ): 623-628.
[ 16 ] FANG H S , XIE S , TAI Y W , et al.RMPE : regional multi-person pose estimation [ C ]// IEEE International Conference on Computer Vision.Italia : Venice , 2017 : 2353-2362.
[ 17 ] 梅阳,王永雄,秦琪,等 . 一种基于关键帧的人体行为识别方法[ J ] . 光学技术, 2017 , 43 ( 4 ): 323-328.
[ 18 ] 管珊珊,张益农 . 基于残差时空图卷积网络的 3D 人体行为识别 [ J ] . 计算机应用与软件,2020 , 37 ( 3 ):198-201 , 250.
[ 19 ] 田联房,吴啟超,杜启亮,等 . 基于人体骨架序列的手扶电梯乘客异常行为识别[ J ] . 华南理工大学学报(自然科学版),2019 , 47 ( 4 ): 10-19.
[ 20 ] BAEK J , KIM J , KIM E.Fast and efficient pedestrian detection via the cascade implementation of an additive kernel support vector machine [ J ] .IEEE Transactions on intelligent transportation systems , 2016 , 18 ( 4 ): 902-916.
[ 21 ] 仉长崎,管业鹏 . 基于动态粒子流场的视频异常行为自动识别[ J ] . 光电子 . 激光, 2015 , 26 ( 12 ): 2375-2380.
[ 22 ] AGYEMAN R , RAFIQ M , SHIN H K , et al.Optimizing spatiotemporal feature learning in 3D convolutional neural networks with pooling blocks [ J ] .IEEE Access , 2021 , 9 : 70797-70805.

备注/Memo

备注/Memo:
收稿日期: 2021-06-19
基金项目:国家自然科学基金资助项目( 61971381 );山西省自然科学基金( 201801D221206 , 201801D221207 );山西省研究生教育创新项目( 2020BY098 )
作者简介:刘嘉宇 ( 1996- ),男,山西太原人,硕士研究生,研究方向为深度学习、行为识别等;陈 平 ( 1983- ),男,山西太原人,博士,教授,研究方向为信号与信息处理、图像处理与重建、人工智能等。
更新日期/Last Update: 2022-02-28