A STUDY ON THE MONITORING MANAGEMENT SYSTEM FOR ANALYZING THE HEALTH STATUS OF LIVESTOCK CATTLE BASED ON MACHINE LEARNING

The current integration of agriculture and IT technology has enhanced the efficiency of farm productivity, maintenance, and management.Particularly in the livestock industry, IT technology is predominantly utilized for farm management.Notably, the use of digital databases allows for the storage and analysis of real-time information regarding livestock, such as estrus detection, genetic status, and movement patterns, facilitating the establishment of comprehensive livestock management systems.These systems for monitoring and analyzing livestock behavior have primarily focused on single functions such as estrus detection, body temperature monitoring, and movement pattern monitoring.Among these, systems analyzing animal behavior have traditionally relied on the use of accelerometers.Such systems operate by aggregating and analyzing the threeaxis values of the accelerometer.However, this paper proposes a multimodal learning approach that independently trains and integrates the values of the three axes, aiming to construct an efficient system for managing cattle and livestock.This approach extends the role of IT technology in the management and monitoring of the livestock industry, enabling more sophisticated analyses and predictions through an innovative application.The multimodal learning approach interprets and integrates the data from each axis separately, providing more precise analyses.This method allows for finer adjustments and improvements in various aspects of livestock management.Such technological advancements hold the potential to significantly enhance the efficiency and productivity in the livestock industry.