基于加速仪运动传感器的牲畜行为监测研究进展
|
摘要
加速仪运动传感器被大量用于牲畜行为监测,是当前开展精准养殖研究的重要方向。综述了加速仪运动传感器在牲畜行为分类方面的研究,介绍了加速仪运动传感器在牲畜身上的可佩戴位置,并对比了其在腿部和颈部的佩戴差异,详细论述了基于加速仪运动传感器的牲畜行为分类流程,具体包括采样时间窗口确定、特征向量抽取、分类算法构建等。行为自动分类是对牲畜进行自动化监测、精准化管理的前期和基础,未来应进一步突破分类模型构建、多元传感器融合、实时数据处理等关键技术研究。
Accelerometer motion sensor is widely used in animal behavior monitoring,which is an important direction of precision breeding research. This paper summarized the research of accelerometer motion sensor in animal behavior classification,introduced the wearable position of the accelerometer motion sensor on the animal body and compared the difference of its wearing in the legs and neck. The process of livestock behavior classification based on accelerometer motion sensor was discussed in detail,including the determination of sampling time window,feature vector extraction and classification algorithm construction. Automatic behavior classification was the preliminary and foundation for automatic monitoring and precision management of livestock. In the future,key technologies such as classification model construction,multi-sensor fusion and real-time data processing should be further broken through.
Accelerometer motion sensor is widely used in animal behavior monitoring,which is an important direction of precision breeding research. This paper summarized the research of accelerometer motion sensor in animal behavior classification,introduced the wearable position of the accelerometer motion sensor on the animal body and compared the difference of its wearing in the legs and neck. The process of livestock behavior classification based on accelerometer motion sensor was discussed in detail,including the determination of sampling time window,feature vector extraction and classification algorithm construction. Automatic behavior classification was the preliminary and foundation for automatic monitoring and precision management of livestock. In the future,key technologies such as classification model construction,multi-sensor fusion and real-time data processing should be further broken through.
引文
[1]沈明霞,刘龙申,闫丽,等.畜禽养殖个体信息监测技术研究进展[J].农业机械学报,2014,45(10):245-251.Shen M X,Liu L S,Yan L,et al.. Review of monitoringtechnology for animal individual in animal husbandry[J].Trans. CSAM,2014,45(10):245-251.
[2] Moreau M,Siebert S,Buerkert A,et al.. Use of a tri-axialaccelerometer for automated recording and classification ofgoats'grazing behavior[J]. Appl. Anim. Behaviour Sci.,2009,119:158-170.
[3] González L A,Bishop-Hurley G J,Handcock R N,et al..Behavioral classification of data from collars containing motionsensors in grazing cattle[J]. Comp. Electron. Agric.,2015,110:91-102.
[4] Daniel S,Rahman A,Greg J,et al.. Behavior classification ofcows fitted with motion collars:Decomposing multi-classclassification into a set of binary problems[J]. Comp.Electron. Agric.,2016,131:40-50.
[5] Arcidiacono C,Porto S M C,Mancino M,et al.. Developmentof a threshold-based classifier for real-time recognition of cowfeeding and standing behavioural activities from accelerometerdata[J]. Comp. Electron. Agric.,2017,134:124-134.
[6] Yoshitoshi R, Watanabe N, Kawamura K, et al..Distinguishing cattle foraging activities using an accelerometry-based activity monitor[J]. Rangeland Ecol. Manage.,2013,66(3):382-386.
[7] Zobel G,Weary D M, Leslie K, et al.. Technical note:Validation of data loggers for recording lying behavior in dairygoats[J]. J. Dairy Sci.,2015,98:1082-1089.
[8] Alvarenga F A P,Borges I,Palkovic L,et al.. Using a three-axis accelerometer to identify and classify sheep behavior atpasture[J]. Appl. Anim. Behaviour Sci.,2016,181:91-99.
[9] Giovanetti V, Decandia M, Molle G, et al.. Automaticclassification system for grazing, ruminating and restingbehavior of dairy sheep using a tri-axial accelerometer[J].Livestock Sci.,2017,196:42-48.
[10] Cornou C,Sren L,Anders R K. Modelling and monitoringsows'activity types in farrowing house using acceleration data[J]. Comp. Electron. Agric.,2011,76:316-324.
[11] Maciej O,Kristina M,Daniel B,et al.. Classification of nest-building behaviour in non-crated farrowing sows on the basis ofaccelerometer data[J]. Biosyst. Engin.,2015,140:48-58.
[12] Liu L S,Shen M X,Yan W,et al.. Acquisition and analysis ofsows'behavior before farrowing based on acceleration sensor[J]. Trans. CSAM,2013,44(3):192-196.
[13] DuBois C,Zakrajsek E,Haley D B,et al.. Validation oftriaxial accelerometers to measure the lying behaviour of adultdomestic horses[J]. Animal,2015,9(1):110-114.
[14] Mc Gowan J E,Burke C R,Jago J G. Validation of a technologyfor objectively measuring behaviour in dairy cows and itsapplication for oestrous detection[J]. NZ Soc. Anim. Prod.,2007,67:136-142.
[15] Awasthi A,Awasthi A,Riordan D,et al.. Non-invasive sensortechnology for the development of a dairy cattle healthmonitoring system[J]. Computers,2016,5(23):1-11.
[16] Rutten C J,Velthuis A G J,Steeneveld W,et al.. Invitedreview:Sensors to support health management on dairy farms[J]. J. Dairy Sci.,2013,96(4):1928-1952.
[17] Helwatkar A,Riordan D, Walsh J. Sensor technology foranimal health monitoring[A]. In:Proceedings of 8thInternational Conference on Sensing Technology[C]. UKLiverpool,2014.
[18] Brad J, Johann F, David G, et al.. Evaluation of two-dimensional accelerometers to monitor behavior of beef calvesafter castration[J]. Am. J. Vet. Res.,2008,69(8):1005-1012.
[19] Descoteaux M,Deriche R,Bihan D,et al.. Embedded sensortechnology for real time determination of animal lying time[J].Comp. Electron. Agric.,2009,66(1):106-111.
[20] Delagarde R,Lamberton P. Daily grazing time of dairy cows isrecorded accurately using the Lifecorder Plus device[J]. Appl.Anim. Behaviour Sci.,2015,165:25-32.
[21] Ledgerwood D N,Winckler C,Tucker C B. Evaluation of dataloggers, sampling intervals, and editing techniques formeasuring the lying behavior of dairy cattle[J]. J. Dairy Sci.,2010,93(11):5129-5139.
[22] Werner J,Leso L,Umstatter C, et al.. Evaluation of theRumiWatchSystem for measuring grazing behaviour of cows[J]. J. Neurosci. Methods,2018,300:138-146.
[23] Bikker J, Laar H V, Rump P, et al.. Technical note:Evaluation of an ear-attached movement sensor to record cowfeeding behavior and activity[J]. J. Dairy Sci.,2014,97(5):2974-2979.
[24] Wolfger B,Timsit E,Pajor E A, et al.. Technical note:Accuracy of an ear tag-attached accelerometer to monitorrumination and feeding behavior in feedlot cattle[J]. J. Anim.Sci.,2015,93:3164-3168.
[25] Watanabe N,Sakanoue S,Kawamura K,et al.. Developmentof an automatic classification system for eating,ruminating andresting behavior of cattle using an accelerometer[J]. GrasslandSci.,2008,54(4):231-237.
[26] Robert B,White B J,Renter D G,et al.. Evaluation of three-dimensional accelerometers to monitor and classify behaviorpatterns in cattle[J]. Comp. Electron. Agric.,2009,67(1):80-84.
[27] Guo L F,Mitchell W,Robin D,et al.. Comparison of grazingbehaviour of sheep on pasture with different sward surfaceheights using an inertial measurement unit sensor[J]. Comp.Electron. Agric.,2018,150:394-401.
[28] Nielsen L R,Pedersen A R,Mette S H,et al.. Quantifyingwalking and standing behaviour of dairy cows using a movingaverage based on output from an accelerometer[J]. Appl.Anim. Behaviour Sci.,2010,127:12-19.
[29] Borchers M R,Chang Y M,Tsai I C,et al.. A validation oftechnologies monitoring dairy cow feeding, ruminating, andlying behaviors[J]. J. Dairy Sci.,2016,99(9):7458-7466.
[30] Martiskainen P, Mikko J, Jukka-Pekka S, et al.. Cowbehaviour pattern recognition using a three-dimensionalaccelerometer and support vector machines[J]. Appl. Anim.Behaviour Sci.,2009,119:32-38.
[31] Vazquez Diosdado J A,Baker Z E, Hodges H R, et al..Classification of behaviour in housed dairy cows using anaccelerometer-based activity monitoring system[J]. Anim.Biotelemetry,2015,3(15):1-14.
[32] Mattachini G, Antler A, Riva E, et al.. Automatedmeasurement of lying behavior for monitoring the comfort andwelfare of lactating dairy cows[J]. Livestock Sci.,2013,158:145-150.
[33] Dutta R, Smith D, Rawnsley R, et al.. Dynamic cattlebehavioural classification using supervised ensemble classifiers[J]. Comp. Electronics Agric.,2015,111:18-28.
[34] Nielsen P P. Automatic registration of grazing behaviour indairy cows using 3D activity loggers[J]. Appl. Anim.Behaviour Sci.,2013,148:179-184.
[35] Arcidiacono C,Porto S M C,Mancino M,et al.. A threshold-based algorithm for the development of inertial sensor-basedsystems to perform real time cow step counting in free-stallbarns[J]. Biosyst. Engin.,2017,153:99-109.
[36] Shen W Z,Chen C C,Zheng S,et al.. The design of systemabout cow activity based on SVM[J]. Int. J. Smart Home,2015,9(3):91-100.
[37] Gutierrez-Galan D, Dominguez-Morales J P, Cerezuela-Escudero E,et al.. Embedded neural network for real-timeanimal behavior classification[J]. Neurocomputing,2018,272:17-26.
[38] Smith D,Dutta R,Hellicar A,et al.. Bag of Class Posteriors,a new multivariate time series classifier applied to animalbehaviour identification[J]. Exp. Syst. Appl.,2015,42:3774-3784.
[39]於少文,孔繁涛,张建华,等.可穿戴设备技术在奶牛养殖中的应用及发展趋势[J].中国农业科技导报,2016,18(5):102-110.Yu S W,Kong F T,Zhang J H,et al.. Application anddevelopment trend of wearable devices technology in dairyfarming[J]. J. Agric. Sci. Technol.,2016,18(5):102-110.
[40]柏广宇,刘龙申,沈明霞,等.基于无线传感器网络的母猪体温实时监测节点研制[J].南京农业大学学报,2014,37(5):128-134.Bai G Y,Liu L S,Shen M X,et al.. Design of sow bodytemperature monitoring node based on wireless sensor network[J]. J. Nanjing Agric. Univ.,2014,37(5):128-134.
[2] Moreau M,Siebert S,Buerkert A,et al.. Use of a tri-axialaccelerometer for automated recording and classification ofgoats'grazing behavior[J]. Appl. Anim. Behaviour Sci.,2009,119:158-170.
[3] González L A,Bishop-Hurley G J,Handcock R N,et al..Behavioral classification of data from collars containing motionsensors in grazing cattle[J]. Comp. Electron. Agric.,2015,110:91-102.
[4] Daniel S,Rahman A,Greg J,et al.. Behavior classification ofcows fitted with motion collars:Decomposing multi-classclassification into a set of binary problems[J]. Comp.Electron. Agric.,2016,131:40-50.
[5] Arcidiacono C,Porto S M C,Mancino M,et al.. Developmentof a threshold-based classifier for real-time recognition of cowfeeding and standing behavioural activities from accelerometerdata[J]. Comp. Electron. Agric.,2017,134:124-134.
[6] Yoshitoshi R, Watanabe N, Kawamura K, et al..Distinguishing cattle foraging activities using an accelerometry-based activity monitor[J]. Rangeland Ecol. Manage.,2013,66(3):382-386.
[7] Zobel G,Weary D M, Leslie K, et al.. Technical note:Validation of data loggers for recording lying behavior in dairygoats[J]. J. Dairy Sci.,2015,98:1082-1089.
[8] Alvarenga F A P,Borges I,Palkovic L,et al.. Using a three-axis accelerometer to identify and classify sheep behavior atpasture[J]. Appl. Anim. Behaviour Sci.,2016,181:91-99.
[9] Giovanetti V, Decandia M, Molle G, et al.. Automaticclassification system for grazing, ruminating and restingbehavior of dairy sheep using a tri-axial accelerometer[J].Livestock Sci.,2017,196:42-48.
[10] Cornou C,Sren L,Anders R K. Modelling and monitoringsows'activity types in farrowing house using acceleration data[J]. Comp. Electron. Agric.,2011,76:316-324.
[11] Maciej O,Kristina M,Daniel B,et al.. Classification of nest-building behaviour in non-crated farrowing sows on the basis ofaccelerometer data[J]. Biosyst. Engin.,2015,140:48-58.
[12] Liu L S,Shen M X,Yan W,et al.. Acquisition and analysis ofsows'behavior before farrowing based on acceleration sensor[J]. Trans. CSAM,2013,44(3):192-196.
[13] DuBois C,Zakrajsek E,Haley D B,et al.. Validation oftriaxial accelerometers to measure the lying behaviour of adultdomestic horses[J]. Animal,2015,9(1):110-114.
[14] Mc Gowan J E,Burke C R,Jago J G. Validation of a technologyfor objectively measuring behaviour in dairy cows and itsapplication for oestrous detection[J]. NZ Soc. Anim. Prod.,2007,67:136-142.
[15] Awasthi A,Awasthi A,Riordan D,et al.. Non-invasive sensortechnology for the development of a dairy cattle healthmonitoring system[J]. Computers,2016,5(23):1-11.
[16] Rutten C J,Velthuis A G J,Steeneveld W,et al.. Invitedreview:Sensors to support health management on dairy farms[J]. J. Dairy Sci.,2013,96(4):1928-1952.
[17] Helwatkar A,Riordan D, Walsh J. Sensor technology foranimal health monitoring[A]. In:Proceedings of 8thInternational Conference on Sensing Technology[C]. UKLiverpool,2014.
[18] Brad J, Johann F, David G, et al.. Evaluation of two-dimensional accelerometers to monitor behavior of beef calvesafter castration[J]. Am. J. Vet. Res.,2008,69(8):1005-1012.
[19] Descoteaux M,Deriche R,Bihan D,et al.. Embedded sensortechnology for real time determination of animal lying time[J].Comp. Electron. Agric.,2009,66(1):106-111.
[20] Delagarde R,Lamberton P. Daily grazing time of dairy cows isrecorded accurately using the Lifecorder Plus device[J]. Appl.Anim. Behaviour Sci.,2015,165:25-32.
[21] Ledgerwood D N,Winckler C,Tucker C B. Evaluation of dataloggers, sampling intervals, and editing techniques formeasuring the lying behavior of dairy cattle[J]. J. Dairy Sci.,2010,93(11):5129-5139.
[22] Werner J,Leso L,Umstatter C, et al.. Evaluation of theRumiWatchSystem for measuring grazing behaviour of cows[J]. J. Neurosci. Methods,2018,300:138-146.
[23] Bikker J, Laar H V, Rump P, et al.. Technical note:Evaluation of an ear-attached movement sensor to record cowfeeding behavior and activity[J]. J. Dairy Sci.,2014,97(5):2974-2979.
[24] Wolfger B,Timsit E,Pajor E A, et al.. Technical note:Accuracy of an ear tag-attached accelerometer to monitorrumination and feeding behavior in feedlot cattle[J]. J. Anim.Sci.,2015,93:3164-3168.
[25] Watanabe N,Sakanoue S,Kawamura K,et al.. Developmentof an automatic classification system for eating,ruminating andresting behavior of cattle using an accelerometer[J]. GrasslandSci.,2008,54(4):231-237.
[26] Robert B,White B J,Renter D G,et al.. Evaluation of three-dimensional accelerometers to monitor and classify behaviorpatterns in cattle[J]. Comp. Electron. Agric.,2009,67(1):80-84.
[27] Guo L F,Mitchell W,Robin D,et al.. Comparison of grazingbehaviour of sheep on pasture with different sward surfaceheights using an inertial measurement unit sensor[J]. Comp.Electron. Agric.,2018,150:394-401.
[28] Nielsen L R,Pedersen A R,Mette S H,et al.. Quantifyingwalking and standing behaviour of dairy cows using a movingaverage based on output from an accelerometer[J]. Appl.Anim. Behaviour Sci.,2010,127:12-19.
[29] Borchers M R,Chang Y M,Tsai I C,et al.. A validation oftechnologies monitoring dairy cow feeding, ruminating, andlying behaviors[J]. J. Dairy Sci.,2016,99(9):7458-7466.
[30] Martiskainen P, Mikko J, Jukka-Pekka S, et al.. Cowbehaviour pattern recognition using a three-dimensionalaccelerometer and support vector machines[J]. Appl. Anim.Behaviour Sci.,2009,119:32-38.
[31] Vazquez Diosdado J A,Baker Z E, Hodges H R, et al..Classification of behaviour in housed dairy cows using anaccelerometer-based activity monitoring system[J]. Anim.Biotelemetry,2015,3(15):1-14.
[32] Mattachini G, Antler A, Riva E, et al.. Automatedmeasurement of lying behavior for monitoring the comfort andwelfare of lactating dairy cows[J]. Livestock Sci.,2013,158:145-150.
[33] Dutta R, Smith D, Rawnsley R, et al.. Dynamic cattlebehavioural classification using supervised ensemble classifiers[J]. Comp. Electronics Agric.,2015,111:18-28.
[34] Nielsen P P. Automatic registration of grazing behaviour indairy cows using 3D activity loggers[J]. Appl. Anim.Behaviour Sci.,2013,148:179-184.
[35] Arcidiacono C,Porto S M C,Mancino M,et al.. A threshold-based algorithm for the development of inertial sensor-basedsystems to perform real time cow step counting in free-stallbarns[J]. Biosyst. Engin.,2017,153:99-109.
[36] Shen W Z,Chen C C,Zheng S,et al.. The design of systemabout cow activity based on SVM[J]. Int. J. Smart Home,2015,9(3):91-100.
[37] Gutierrez-Galan D, Dominguez-Morales J P, Cerezuela-Escudero E,et al.. Embedded neural network for real-timeanimal behavior classification[J]. Neurocomputing,2018,272:17-26.
[38] Smith D,Dutta R,Hellicar A,et al.. Bag of Class Posteriors,a new multivariate time series classifier applied to animalbehaviour identification[J]. Exp. Syst. Appl.,2015,42:3774-3784.
[39]於少文,孔繁涛,张建华,等.可穿戴设备技术在奶牛养殖中的应用及发展趋势[J].中国农业科技导报,2016,18(5):102-110.Yu S W,Kong F T,Zhang J H,et al.. Application anddevelopment trend of wearable devices technology in dairyfarming[J]. J. Agric. Sci. Technol.,2016,18(5):102-110.
[40]柏广宇,刘龙申,沈明霞,等.基于无线传感器网络的母猪体温实时监测节点研制[J].南京农业大学学报,2014,37(5):128-134.Bai G Y,Liu L S,Shen M X,et al.. Design of sow bodytemperature monitoring node based on wireless sensor network[J]. J. Nanjing Agric. Univ.,2014,37(5):128-134.