This study focuses on the design and implementation of an automated system for monitoring the activities of rabbits by combining Internet of Things technology with a Support Vector Machine classification model. The system employs a micro controller and an inertial measurement sensor to continuously capture motion data from rabbits in real time. These motion readings are transmitted to a database, where they are processed and analyzed to identify specific behavioral categories, namely feeding, moving, and resting. The classification process relies on a machine learning approach capable of handling complex and non-linear patterns in the data, enabling more accurate recognition of subtle differences between activities.
The motivation behind this development is to address the inefficiencies and inaccuracies of traditional livestock monitoring methods, which often depend on manual observation and are prone to human error. By automating data collection and classification, the system aims to improve both the precision and efficiency of livestock management, allowing for early detection of abnormal behavior and better overall animal welfare.
Initial testing of the classification model achieved an accuracy rate of 36 percent, which significantly improved to 84 percent after parameter optimization. This substantial increase highlights the importance of fine-tuning model parameters to enhance performance. The results demonstrate that integrating sensor-based monitoring with advanced classification techniques offers a practical and effective solution for modernizing livestock surveillance, paving the way for more intelligent and data-driven agricultural practices.