Authors:

Neelesh Sarathy

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Monitoring pig behavior effectively is essential for ensuring animal well-being and improving farm operations. Traditional manual observation methods are labor intensive and prone to error. Recent advancements in computer vision and deep learning offer opportunities to automate behavior recognition enhancing efficiency and accuracy in managing livestock. This study aims to create an automated system for identifying pig behaviors such as resting, eating, drinking and aggressive actions using a YOLOv8 based model trained on video data of nursery pigs. Video data was gathered from cameras mounted on the ceiling of pig pens. YOLOv8, known for its object detection capabilities, was used to train the model to detect pigs and recognize their behaviors. Preprocessing of data involved annotating and enhancing videos to train the model. The model is projected to achieve an accuracy of 95 percent. This suggests that automating behavior recognition can greatly improve pig welfare management by providing data. Future research will concentrate on automatic monitoring of farrowing pigs and refining the model for increased accuracy and expanding its use to monitor animal behaviors across species. Keywords: Pigs; Machine Learning; Animal Monitoring

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Neelesh Sarathy