基于高斯混合-隐马尔科夫融合算法识别奶牛步态时相

doi:10.12133/j.smartag.SA202204003

摘要/Abstract

摘要：

奶牛步态时相是反映奶牛健康及跛行严重程度的重要指标。为准确自动识别奶牛步态时相，本研究提出一种融合高斯混合模型（Gaussian Mixture Model，GMM）和隐马尔科夫模型（Hidden Markov Model，HMM）的无监督学习奶牛步态时相识别算法 GMM-HMM。使用惯性测量单元采集奶牛后肢加速度和角速度信号，通过卡尔曼滤波消除噪声，筛选并提取特征值，构建GMM-HMM模型，实现奶牛静立相、连续步态中的站立相和摆动相等3种步态时相的自动识别。结果表明，静立相识别的准确率、召回率和F₁分别为89.28%、90.95%和90.91%，连续步态中的站立相识别的准确率、召回率和F₁分别为91.55%、86.71%和89.06%，连续步态中的摆动相识别的准确率、召回率和F₁分别为86.67%、91.51%和89.03%。奶牛步态分割的准确率为91.67%，相较于基于事件的峰值检测法和动态时间规整算法准确率分别提高了4.23%和1.1%。本研究可为下一步基于穿戴式步态分析的奶牛跛行特征提取提供技术参考。

关键词: 奶牛跛行, 步态时相, 步态分割, 高斯混合模型, 隐马尔科夫模型, 卡尔曼滤波

Abstract:

The gait phase of dairy cows is an important indicator to reflect the severity of lameness. IThe accuracy of available gait segmentation methods was not enough for lameness detection. In this study, a gait phase recognition method based on Gaussian mixture model (GMM) and hidden Markov model (HMM) was proposed and tested. Firstly, wearable inertial sensors LPMS-B2 were used to collect the acceleration and angular velocity signals of cow hind limbs. In order to remove the noise of the system and restore the real dynamic data, Kalman filter was used for data preprocessing. The first-order difference of the angular velocity of the coronal axis was selected as the eigenvalue. Secondly, to analyze the long-term continuous recorded gait sequences of dairy cows, the processed data was clustered by GMM in the unsupervised way. The clustering results were taken as the input of the HMM, and the gait phase recognition of dairy cows was realized by decoding the observed data. Finally, the cow gait was segmented into 3 phases, including the stationary phase, standing phase and swing phase. At the same time, gait segmentation was achieved according to the standing phase and swing phase. The accuracy, recall rate and F₁ of the stationary phase were 89.28%, 90.95% and 90.91%, respectively. The accuracy, recall rate and F₁ of the standing phase recognition in continuous gait were 91.55%, 86.71% and 89.06%, respectively. The accuracy, recall rate and F₁ of the swing phase recognition in continuous gait were 86.67%, 91.51% and 89.03%, respectively. The accuracy of cow gait segmentation was 91.67%, which was 4.23% and 1.1 % higher than that of the event-based peak detection method and dynamic time warping algorithm, respectively. The experimental results showed that the proposed method could overcome the influence of the cow's walking speed on gait phase recognition results, and recognize the gait phase accurately. This experiment provides a new method for the adaptive recognition of the cow gait phase in unconstrained environments. The degree of lameness of dairy cows can be judged by the gait features.

Key words: dairy cow lameness, gait phase, gait segmentation, gaussian mixture model, hidden Markov model, Kalman filtering

中图分类号:

张楷, 韩书庆, 程国栋, 吴赛赛, 刘继芳. 基于高斯混合-隐马尔科夫融合算法识别奶牛步态时相[J]. 智慧农业(中英文), 2022, 4(2): 53-63.

ZHANG Kai, HAN Shuqing, CHENG Guodong, WU Saisai, LIU Jifang. Gait Phase Recognition of Dairy Cows based on Gaussian Mixture Model and Hidden Markov Model[J]. Smart Agriculture, 2022, 4(2): 53-63.

图/表 11

图1

图2

图3

图4

图5

图6

图7

图8

表1

图9

表2

参考文献 20

1	TADICH N, FLOR E, GREEN L. Associations between hoof lesions and locomotion score in 1098 unsound dairy cows[J]. The Veterinary Journal, 2010, 184(1): 60-65.
2	OEHM A W, KNUBBEN-SCHWEIZER G, RIEGER A, et al. A systematic review and meta-analyses of risk factors associated with lameness in dairy cows[J]. BMC Veterinary Research, 2019, 15(1): ID 346.
3	邵军, 林为民, 孙新文, 等. 石河子地区奶牛肢蹄病类型和发病情况调查分析[J]. 中国奶牛, 2017(1): 30-33.
	SHAO J, LIN W, SUN X, et al. Investigation and analysis of limb hoof disease types and incidence in dairy cows in Shihezi area[J]. China Dairy Cattle, 2017 (1) : 30-33.
4	韩书庆, 张晶, 程国栋, 等. 奶牛跛行自动识别技术研究现状与挑战[J]. 智慧农业(中英文), 2020, 2(3): 21-36.
	HAN S, ZHANG J, CHENG G, et al. Research status and challenges of automatic identification technology for cow lameness[J]. Smart Agriculture, 2020, 2 (3) : 21-36.
5	程国栋, 吴建寨, 邢丽玮, 等. 基于IMU的细粒度奶牛行为判别[J]. 中国农业大学学报, 2022, 27(4):179-186.
	CHENG G, WU J, XING L, etc. The behavior discrimination of fine-grained dairy cows based on IMU[J].Journal of China Agricultural University, 2022, 27 (4):179-186.
6	LI Z, ZHANG Q, LYU S, et al. Fusion of RGB, optical flow and skeleton features for the detection of lameness in dairy cows[J]. Biosystems Engineering, 2022, 218: 62-77.
7	ALSAAOD M, NIEDERHAUSER J J, BEER G, et al. Development and validation of a novel pedometer algorithm to quantify extended characteristics of the locomotor behavior of dairy cows[J]. Journal of Dairy Science, 2015, 98(9): 6236-6242.
8	ZHAO K, BEWLEY J M, HE D, et al. Automatic lameness detection in dairy cattle based on leg swing analysis with an image processing technique[J]. Computers and Electronics in Agriculture, 2018, 148: 226-236.
9	宋怀波, 姜波, 吴倩, 等. 基于头颈部轮廓拟合直线斜率特征的奶牛跛行检测方法[J]. 农业工程学报, 2018, 34(15): 190-199.
	SONG H, JIANG B, WU Q, et al. Detection method of cow lameness based on fitting straight line slope characteristics of head and neck contour[J]. Transactions of the CASE, 2018, 34 (15) : 190-199.
10	JIANG B, SONG H, HE D. Lameness detection of dairy cows based on a double normal background statistical model[J]. Computers and Electronics in Agriculture, 2019, 158: 140-149.
11	PENG Y, KONDO N, FUJIURA T, et al. Classification of multiple cattle behavior patterns using a recurrent neural network with long short-term memory and inertial measurement units[J]. Computers and Electronics in Agriculture, 2019, 157: 247-253.
12	HALADJIAN J, BRÜGGE B, NÜSKE S. An approach for early lameness detection in dairy cattle[C]// The 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and The 2017 ACM International Symposium on Wearable Computers. New York, USA: Association for Computing Machinery, 2017.
13	苏力德, 张永, 王健, 等. 基于改进动态时间规整算法的奶牛步态分割方法[J]. 农业机械学报, 2020, 51(7): 52-59.
	SU L, ZHANG Y, WANG J, et al. Gait segmentation method for dairy cows based on improved dynamic time warping algorithm[J]. Transactions of the CSAM, 2020, 51 (7) : 52-59.
14	张向刚, 唐海, 付常君, 等. 一种基于隐马尔科夫模型的步态识别算法[J]. 计算机科学, 2016, 43(7): 285-289, 302.
	ZHANG X, TANG H, FU C, et al. A gait recognition algorithm based on hidden Markov model[J]. Computer Science, 2016, 43 (7) : 285-289, 302.
15	HAJI GHASSEMI N, HANNINK J, MARTINDALE C F, et al. Segmentation of gait sequences in sensor-based movement analysis: A comparison of methods in Parkinson's disease[J]. Sensors (Basel), 2018, 18(1):ID 145
16	ROTH N, KUDERLE A, ULLRICH M, et al. Hidden Markov model based stride segmentation on unsupervised free-living gait data in Parkinson's disease patients[J]. Journal of Neuro Engineering and Rehabilitation, 2021, 18(1): ID 93.
17	王东源, 王玉舜, 谷禹, 等. 天津地区奶牛蹄病分析研究[J]. 天津农学院学报, 2015, 22(2): 37-39.
	WANG D, WANG Y, GU Y, et al. Analysis of cow hoof disease in Tianjin[J]. Journal of Tianjin Agricultural University, 2015, 22(2): 37-39.
18	XIA R, ZHANG Q, DENG X. Multiscale Gaussian convolution algorithm for estimate of Gaussian mixture model[J]. Communications in Statistics-Theory and Methods, 2018, 48(23): 5889-5910.
19	RABINER L R. A Tutorial on Hidden Markov models and selected applications in speech recognition[M]// WAIBEL A, LEE K-F. Readings in speech recognition. San Francisco: Morgan Kaufmann, 1990.
20	ALSAAOD M, LUTERNAUER M, HAUSEGGER T, et al. The cow pedogram—Analysis of gait cycle variables allows the detection of lameness and foot pathologies[J]. Journal of Dairy Science, 2017, 100(2): 1417-1426.

时相	初始概率	权重	均值	方差
静立相	0	1	0.002	0.027
站立相	0	1	-0.251	3.904
摆动相	1	1	0.357	1976.724

[1]	郭大方, 杜岳峰, 武秀恒, 侯思余, 栗晓宇, 张延安, 陈度. 农机装备数字孪生：从概念到应用[J]. 智慧农业(中英文), 2023, 5(2): 149-160.
[2]	宋淑然, 胡圣洋, 孙道宗, 代秋芳, 薛秀云, 谢家兴, 李震. 山地果园管道自动喷雾系统设计与试验[J]. 智慧农业(中英文), 2022, 4(3): 86-94.
[3]	王辉, 陈睿鹏, 余志雪, 贺越, 张帆, 熊本海. 基于卟啉和半导体单壁碳纳米管的场效应气体传感器检测草莓恶疫霉[J]. 智慧农业(中英文), 2022, 4(3): 143-151.
[4]	郭秀明, 诸叶平, 李世娟, 张杰, 吕纯阳, 刘升平. 农业复杂环境下尺度自适应小目标识别算法——以蜜蜂为研究对象[J]. 智慧农业(中英文), 2022, 4(1): 140-149.
[5]	王子杰, 刘国海, 张多, 沈跃, 姚震, 张贺. 高地隙四轮独立驱动喷雾机路径跟踪模型预测控制[J]. 智慧农业(中英文), 2021, 3(3): 82-93.
[6]	王芳, 滕桂法, 姚竟发. 带时间窗的多目标蔬菜运输配送路径优化算法[J]. 智慧农业(中英文), 2021, 3(3): 152-161.
[7]	张小青, 邵松, 郭新宇, 樊江川. 田间玉米苗期高通量动态监测方法[J]. 智慧农业（中英文）, 2021, 3(2): 88-99.
[8]	郝王丽, 尉培岩, 郝飞, 韩猛, 韩冀皖, 孙玮蓉, 李富忠. 基于YOLOv4和自适应锚框调整的谷穗检测方法[J]. 智慧农业(中英文), 2021, 3(1): 63-74.
[9]	朱超, 吴凡, 刘长斌, 赵健翔, 林丽丽, 田雪莹, 苗腾. 基于超体素聚类和局部特征的玉米植株点云雄穗分割[J]. 智慧农业(中英文), 2021, 3(1): 75-85.
[10]	李燕, 沈杰, 谢航, 高广垠, 刘建雄, 刘洁. 基于轮廓坐标系转换拟合的柚子果形检测分级方法[J]. 智慧农业(中英文), 2021, 3(1): 86-95.
[11]	毛文菊, 刘恒, 王东飞, 杨福增, 刘志杰. 面向果园多机器人通信的AODV路由协议改进设计与测试[J]. 智慧农业(中英文), 2021, 3(1): 96-108.
[12]	朱登胜, 方慧, 胡韶明, 王文权, 周延锁, 王红艳, 刘飞, 何勇. 农机远程智能管理平台研发及其应用[J]. 智慧农业（中英文）, 2020, 2(2): 67-81.
[13]	周成全, 叶宏宝, 俞国红, 胡俊, 徐志福. 基于机器视觉与深度学习的西兰花表型快速提取方法研究[J]. 智慧农业（中英文）, 2020, 2(1): 121-132.
[14]	万亮, 岑海燕, 朱姜蓬, 张佳菲, 杜晓月, 何勇. 基于纹理特征与植被指数融合的水稻含水量无人机遥感监测[J]. 智慧农业（中英文）, 2020, 2(1): 58-67.
[15]	张海峰, 李杨, 张宇, 宋丽娟, 唐立新, 毕洪文. 基于云服务的棚室蔬菜智能终端系统设计与实现——以黑龙江省为研究案例[J]. 智慧农业, 2019, 1(3): 87-99.