Real-Time Behavioral Recognition in Dairy Cows Based on Geomagnetism and Acceleration Information

Abstract

The feeding, ruminating(standing, lying down), running, being still (standing, lying down), head-shaking, drinking, and walking behaviors of dairy cows can reflect their health status. In this study, a multi-sensor was used to collect data of cow's multi-behaviors for research on behavior recognition. Firstly, a collar style data acquisition system was designed using geomagnetic and acceleration sensors to collect the behavioral data of dairy cows during their daily activities. Secondly, the dairy cow behavioral recognition fusion model based on K-Nearest-Neighbors (KNN) and Random Forest (RF) models were used for behavior classification. To verify the accuracy of the fusion model, the algorithms of KNN, RF, Gradient Boosting Decision Tree (GBDT), Support Vector Machine (SVM), and Learning Vector Quantization (LVQ) were introduced for comparative recognition experiments among different algorithms. The KNN-RF fusion model had the highest average recognition accuracy of 98.51%, followed by the KNN model with an average recognition accuracy of 95.37%, and the LVQ model had the lowest average recognition accuracy of 80.81%. For the recognition and verification of each behavior, the KNN-RF fusion model had the most obvious improvement in the recognition of dairy cow feeding behavior, with a recognition accuracy of 99.34%, followed by the KNN model with a recognition accuracy of 95.07%. All six models had the lowest recognition accuracy for cow head-shaking behavior: a recognition accuracy of 89.11% with the KNN-RF model followed by the RF model with a recognition accuracy of 85.14%. The system can quickly and continuously collect cow behavior information, accurately recognize individual behaviors, and provide a scientific basis for the optimal design and efficient management of digital facilities and equipment for dairy cows.

CNN and transfer learning-based classification model for automated cow’s feeding behaviour recognition from accelerometer data

“…The copyright holder for this preprint this version posted July 5, 2022. ; https://doi.org/10.1101/2022.07.03.498612 doi: bioRxiv preprint with the performance reviewed in Table A1 were those by Arcidiacono et al (2017) with F1=93.3%, by Tian et al (2021) with F1= 98.51%, and by Riaboff et al (2020) with a total accuracy of 98%. Exact comparison of the accuracy scores is impractical because of differences in the research conditions, such as experiment environment, sensors and amount and type of the collected data.…”

“…The F1 score achieved in this study was high compared to the systems using NN for cow behaviour classification such as those by Pavlovic et al (2020) with F1=82%, Peng et al (2019) with F1=88.7% and Li, C. et al (2021) with F1=94.4%. Among the systems using machine learning with the performance reviewed in Table A1 were those by Arcidiacono et al (2017) with F1=93.3%, by Tian et al (2021) with F1= 98.51%, and by Riaboff et al (2020) with a total accuracy of 98%. Exact comparison of the accuracy scores is impractical because of differences in the research conditions, such as experiment environment, sensors and amount and type of the collected data.…”

“…Table A1 were those byArcidiacono et al (2017) with F1=93.3%, byTian et al (2021) with F1= 98.51%, and byRiaboff et al (2020) with a total accuracy of 98%. Exact comparison of the accuracy scores is impractical because of differences in the research conditions, such as experiment environment, sensors and amount and type of the collected data.…”

A New Dataset for Video-based Cow Behavior Recognition

“…The fast pathway also used a ResNet3D network with a depth of 50, without lateral connections, and 8 channels. The kernel size of the first convolutional layer was (5,7,7), and the first pooling layer had a stride of 1 and a spatial stride of (1, 2, 2, 1).…”

Cow Behavioural Activities in Extensive Farms: Challenges of Adopting Automatic Monitoring Systems

“…Threshold methods [46,53] Statistical models -Logistic Regression (LR), Hidden Markov Models (HMM), Linear Mixed Models [38,54] Machine learning Supervised Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Support Vector Machine (SVM), k-Nearest Neighbour (k-NN), Naïve Bayes models (NB), Decision Trees (DT) [55,56] Unsupervised k-means [57] Ensemble Random Forest (RF), Extreme Gradient Boosting (XGB), Adaboost (ADA) [24,58] Deep learning…”