Serviços Personalizados
Journal
Artigo
Indicadores
- Citado por SciELO
- Acessos
Links relacionados
- Citado por Google
- Similares em SciELO
- Similares em Google
Compartilhar
Biotecnología en el Sector Agropecuario y Agroindustrial
versão impressa ISSN 1692-3561
Resumo
RAMIREZ-AGUDELO, JOHN-FREDY,; BEDOYA-MAZO, SEBASTIAN,; POSADA-OCHOA, SANDRA-LUCIA, e ROSERO-NOGUERA, JAIME-RICARDO,. Automatic cattle activity recognition on grazing systems. Rev.Bio.Agro [online]. 2022, vol.20, n.2, pp.117-128. Epub 01-Jul-2022. ISSN 1692-3561. https://doi.org/10.18684/rbsaa.v20.n2.2022.1940.
The use of collars, pedometers or activity tags for cattle behavior recognition in short periods (e.g. 24 h) is expensive. Under this particular situation, the development of low-cost and easy-to-use technologies is relevant. Similar to smartphone apps for human activity recognition, which analyzes data from embedded triaxial accelerometer sensors, we develop an Android app to record activity in cattle. Four main steps were followed: a) data acquisition for model training, b) model training, c) app deploy, and d) app utilization. For data acquisition, we developed a system in which three components were used: two smartphones and a Google Firebase account for data storage. A dataset with 945415 rows and four columns was made. Each row corresponds to the three-axis accelerometer values and the cattle activities (Eating, Resting, or Rumination). For model training, the generated database was used to train a recurrent neural network. The performance of the training was assessed by a confusion matrix. For all actual activities, the trained model provided a high prediction (> 96 %). The trained model was used to deploy an Android app by using the TensorFlow API. Finally, three cell phones (LG gm730) were used to test the app and record the activity of six Holstein cows (3 lactating and 3 non-lactating). Direct and non-systematic observations of the animals were made to evaluate the performance of the app. After training, the model’s accuracy was 100, 99, and 96 % for Eating, Rumination, and Resting, respectively. Our results show consistency between the direct observations and the activity recorded by our Android app. In conclusion, this work shows that it is possible to develop low-cost technologies to record the daily activity of grazing cows using the smartphone acceleration data analysis.
Palavras-chave : Android App; Accelerometers; Tensorflow; Animal Behavior; Precision Livestock Farming.