Research on Remote Monitoring Model of Livestock Health Based on Computer Deep Learning

• Published in: 2024 IEEE 2nd International Conference on Image Processing and Computer Applications (ICIPCA) pp. 565 - 569
• Main Author: Xu, Xiaoqing
• Format: Conference Proceeding
• Language: English
• Published: IEEE 28.06.2024
• Subjects: Deep learning; Farming; Image recognition; intelligent farming; Internet of Things; Livestock; Passwords; Real-time systems; Remote monitoring; Sockets; STM32; virtual simulation; Visualization; Wireless LAN
• DOI: 10.1109/ICIPCA61593.2024.10709243

A livestock individual status information acquisition device is established and integrated with a visual virtual simulation application. The serial port transmission completes the work of transmitting to the MT7688 system. In this paper, a wireless LAN based on 4 GEC204G is proposed, which realizes real-time transmission to 4G network by Socket. The identification of animals is obtained through personal identification of animal images on the cloud platform. The animal ID was obtained from the image data through the individual recognition model, and the health indicator information was evaluated to obtain the health status of the livestock. VGGNet-16 and ResNet-50 are used to learn ImageNet respectively, and then they are adjusted according to the selected samples, and finally the model is fused by means method. Remote access to the database was obtained with the authorization of the user account and password. The data collected by the sensor is viewed and analyzed and processed on the client page. This study provides a useful basis for regulating the feeding environment, monitoring the movement state and health status of livestock.