基于数字画像的人体运动信息特征采集方法

doi:10.13229/j.cnki.jdxbgxb.20240642

摘要/Abstract

摘要：

在研究人体运动过程中，为进一步解决现阶段采集方法存在的采集效果差、采集精度低和采集执行时间长等问题，提出了一种基于数字画像的人体运动信息特征采集方法。首先，利用数字画像技术获取人体运动视频图像的标志点，并以标志点为中心，采用改进背景差分算法提取人体运动轮廓。其次，采用外极线约束法与灰度互相关算法处理运动轮廓，得到人体运动轨迹。最后，采用动态时间变形算法得到轨迹的运动类型、速度等相关信息，完成人体运动信息特征的采集。随机选取一组人体运动图像数据对本文方法进行验证，实验结果表明，本文方法的信息特征采集效果好、采集精度高、采集执行时间少。基于数字画像的人体运动信息特征采集方法可以有效获取人体运动信息，实现人体运动信息特征采集，对提高人体运动信息特征的采集质量具有重要意义。

关键词: 运动标志点, 运动目标轮廓, 背景差分, 获取运动轨迹, 动态时间变形法

Abstract:

The collection of information features is an indispensable step in the process of studying human motion. In order to further optimize the problems existing in the current collection methods， such as poor collection effect， low collection accuracy， and long collection execution time， this paper proposes a method of human motion information feature collection based on digital portrait. Firstly， this method uses digital portrait technology to obtain landmarks of human motion video image， and uses improved background difference algorithm to extract human motion contour with landmarks as the center. Secondly， the epipolar constraint method and gray cross correlation algorithm are used to process the motion contour， and the human motion trajectory is obtained. Finally， the dynamic time warping algorithm is used to obtain the motion type， speed and other related information of the trajectory， and complete the collection of human motion information features. A group of human motion image data was randomly selected to verify the proposed method. The experimental results show that the proposed method has good information feature acquisition effect， high acquisition accuracy， and low acquisition execution time. The method of human motion information feature collection based on digital portrait can effectively obtain human motion information and realize human motion information feature collection， which is of great significance for improving the quality of human motion information feature collection.

Key words: sports marker, extract moving target contour, background difference, motion track, dynamic time warping

中图分类号:

R319

苗慧,李明. 基于数字画像的人体运动信息特征采集方法[J]. 吉林大学学报(工学版), 2024, 54(12): 3637-3645.

Hui MIAO,Ming LI. Method of collecting human motion information features based on digital portraits[J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(12): 3637-3645.

图/表 7

图1

图2

图3

表1

实验参数设置"

参数序号	参数名称	参数值
01	RGB三种颜色的敏感加权系数	3、4、2
02	模板的大小 $p$ 、 $q$	6、6
03	高精度像素分辨率 $a$ 、 $b$	3 840×2 160像素
04	动作未着地	0
05	动作着地	1

表1

图4

表2

表3

参考文献 19

1	Sun Y, Wang S, Du X, et al. Skin-conformal MXene-doped wearable sensors with self-adhesive, dual-mode sensing, and high sensitivity for human motions and wireless monitoring[J]. Journal of Materials Chemistry B, 2021, 9(41):8667-8675.
2	徐博,蒲东兵,王一可,等. 基于E-DCPose的视频人体姿态估计[J].东北师大学报: 自然科学版,2024,56(4): 53-61.
	Xu Bo, Pu Dong-bing, Wang Yi-ke,et al. Video human pose estimation based on E-DCPose[J]. Journal of Northeast Normal University (Natural Science Edition), 2024,56(4): 53-61.
3	张震, 李浩方, 李孟洲, 等.改进YOLOv3算法与人体信息数据融合的视频监控检测方法[J]. 郑州大学学报: 工学版, 2021, 42(1): 28-34.
	Zhang Zhen, Li Hao-fang, Li Meng-zhou, et al. Improved YOLOv3 algorithm and human body information data fusion video surveillance detection method[J]. Journal of Zhengzhou University (Engineering and Technology Edition), 2021,42(1):28-34.
4	李文平, 袁强, 陈璐, 等. 基于雷达与图像数据融合的人体目标检测方法[J]. 数据采集与处理, 2021, 36(2): 324-333.
	Li Wen-ping, Yuan Qiang, Chen Lu, et al. Human target detection method based on radar and image data fusion[J]. Data Acquisition and Processing, 2021,36(2): 324-333.
5	郝占军, 党建武, 张岱阳. 一种基于信道状态信息的高鲁棒性人体动作识别方法[J]. 兰州交通大学学报, 2020, 39(6): 37-44.
	Hao Zhan-jun, Dang Jian-wu, Zhang Dai-yang. A robust human action recognition method based on channel state information[J]. Journal of Lanzhou Jiaotong University, 2020,39(6):37-44.
6	郭毅博, 孟文化, 范一鸣, 等. 基于可穿戴传感器数据的人体行为识别数据特征提取方法[J]. 计算机辅助设计与图形学学报, 2021, 33(8): 1246-1253.
	Guo Yi-bo, Meng Wen-hua, Fan Yi-ming, et al. Feature extraction method of human behavior recognition data based on wearable sensor data[J]. Journal of Computer-Aided Design and Graphics, 2021,33(8):1246-1253.
7	胡春生, 赵汇东. 基于腕部IMU数据采集的人体运动模式识别[J]. 宁夏大学学报: 自然科学版, 2021, 42(1): 45-50, 57.
	Hu Chun-sheng, Zhao Hui-dong. Human motion pattern recognition based on wrist IMU data acquisition [J]. Journal of Ningxia University (Natural Science Edition), 2021,42(1):45-50, 57.
8	张宗华, 王森, 王宇莹, 等. 基于编码环带径向直线拟合的圆环编码标志点中心提取方法[J]. 激光与光电子学进展, 2020, 57(7): 212-220.
	Zhang Zong-hua, Wang Sen, Wang Yu-ying, et al. Extraction method of ring coded mark point center based on radial straight line fitting[J]. Advances in Laser and Optoelectronics, 2020,57(7):212-220.
9	班朝, 任国营, 王斌锐, 等. 融合加权SVD算法的工业机器人运动学标定[J]. 计量学报, 2021, 42(9): 1128-1135.
	Ban Chao, Ren Guo-ying, Wang Bin-rui, et al. Kinematic calibration of industrial robot based on fusion weighted SVD algorithm[J]. Acta Metrologica Sinica,2021,42(9):1128-1135.
10	方尔正, 孙纯, 桂晨阳. 矢量舷侧阵平台自噪声空域滤波抑制方法[J]. 哈尔滨工程大学学报, 2020, 41(11): 1636-1641.
	Fang Er-zheng, Sun Chun, Gui Chen-yang. Spatial filtering method for self-noise suppression of vector side array platform[J]. Journal of Harbin Engineering University, 2020,41(11):1636-1641.
11	赵传, 郭海涛, 王优扬, 等. 利用邻域方向分布的机载激光雷达点云建筑物外轮廓提取[J].光学精密工程, 2021, 29(2): 374-387.
	Zhao Chuan, Guo Hai-tao, Wang You-yang, et al. Building exterior contour extraction from airborne laser radar point cloud using neighborhood direction distribution[J]. Optics and Precision Engineering, 2021, 29(2): 374-387.
12	宋雨浓, 林舜江, 唐智强, 等. 基于动态车流的电动汽车充电负荷时空分布概率建模[J]. 电力系统自动化, 2020, 44(23): 47-56.
	Song Yu-nong, Lin Shun-jiang, Tang Zhi-qiang, et al. Probabilistic modeling of spatial-temporal distribution of electric vehicle charging load based on dynamic traffic flow[J]. Automation of Electric Power Systems, 2020,44(23):47-56.
13	宋丽梅, 吕昆昆, 杨燕罡. 融合深度学习与外极线约束的三维人体位姿测量方法[J]. 应用光学, 2020, 41(6): 1166-1173.
	Song Li-mei, Kun-kun Lyu, Yang Yan-gang. 3D Human pose measurement method combining deep learning and outer pole constraint[J]. Journal of Applied Optics, 2020,41(6):1166-1173.
14	郑新涛, 苏道玉. 基于小波矩的智能手绘草图识别系统设计[J]. 现代电子技术, 2021, 44(12): 177-181.
	Zheng Xin-tao, Su Dao-yu. Design of intelligent hand-drawn sketch recognition system based on wavelet moment[J]. Modern Electronic Technique,2021,44(12): 177-181.
15	谢迎娟, 许海燕, 李丽媛, 等.物体坐标系曲面投影立体环视图像生成算法[J]. 小型微型计算机系统, 2022, 43(1): 56-60.
	Xie Ying-juan, Xu Hai-yan, Li Li-yuan, et al. An algorithm for generating stereoscopic circumnavigation image with surface projection in object coordinate system[J]. Small Microcomputer Systems,2022,43(1):56-60.
16	吴晓平, 祁亨年. 基于延迟时间测量的移动源运动参数估计方法[J].传感技术学报, 2020, 33(6): 895-901.
	Wu Xiao-ping, Qi Heng-nian. Estimation method of moving source motion parameters based on delay time measurement[J]. Journal of Sensor Technology, 2020, 33(6): 895-901.
17	丁飞, 朱跃, 艾成万, 等. 基于多模型融合的室内行人航迹推算建模与性能分析[J]. 江苏大学学报: 自然科学版, 2024, 45(4): 456-463.
	Ding Fei, Zhu Yue, Ai Cheng-wan,et al. Modeling and performance analysis of indoor PDR based on multi-model fusion[J]. Journal of Jiangsu University(Natural Science Edition), 2024, 45(4): 456-463.
18	田亮, 金积德, 郑庆祥. 基于检测增强型YOLOv3-tiny的道路场景行人检测[J]. 江苏大学学报: 自然科学版, 2024, 45(4): 441-448.
	Tian Liang, Jin Ji-de, Zheng Qing-xiang. Road scene pedestrian detection based on detection-enhanced YOLOv3-tiny[J]. Journal of Jiangsu University (Natural Science Edition), 2024, 45(4): 441-448.
19	殷炜宏, 王若愚, 段倩倩, 等. 基于时态边缘算子的时间序列自主分段表示法[J]. 计算机工程与科学,2021, 43(6): 1104-1111.
	Yin Wei-hong, Wang Ruo-yu, Duan Qian-qian, et al. Autonomous segmentation representation of time series based on temporal edge operators[J]. Computer Engineering and Science,2021,43(6): 1104-1111.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

实验次数/组	准确率/%
实验次数/组	本文方法	文献［3］方法	文献［4］方法
1	96	32	51
2	92	35	52
3	95	34	51
4	96	34	42
5	92	28	56
6	93	34	53
7	92	36	52
8	93	35	52
9	95	33	53
10 11 12 13 14 15 16 17 18 19 20	97 91 94 93 91 92 93 97 95 96 92	38 40 28 39 40 37 36 37 41 29 37	53 45 56 60 58 39 54 57 39 53 55

图像数量/张	采集执行时间/s
图像数量/张	本文方法	文献［3］方法	文献［4］方法
50	1.21	2.91	3.11
100	1.17	3.52	3.23
150	1.20	3.18	3.25
200	1.21	3.39	3.16
250	1.15	3.53	3.27
300	1.19	3.65	3.29
350	1.17	3.75	3.17
400	1.17	3.86	3.15
450	1.19	3.87	3.16
500	1.20	3.73	3.22

[1]	李志慧, 夏英集, 曲昭伟, 任景琛. 视频监控的数据驱动背景模型[J]. 吉林大学学报(工学版), 2017, 47(4): 1286-1294.
[2]	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.
[3]	陈大伟, 刘海龙, 李金屏. 复杂静态背景下多移动目标实时检测系统的FPGA实现[J]. 吉林大学学报(工学版), 2013, 43(增刊1): 287-290.