Research on a deep learning-based epidemic surveillance system for live poultry transport

Disease surveillance in live poultry transport has an important impact on the prevention and control of live poultry diseases. Traditional poultry disease monitoring in transport mainly relies on the driver's visual observation and experience, which cannot monitor the health status of live poultry in time and relies heavily on the driver's transport experience, resulting in high mortality and risk of disease transmission during transport. A deep learning model is used to establish a live poultry transport epidemic monitoring system, which can achieve real-time photography of the health status of live poultry in transit and automatic comparison, giving full play to the advantages of deep learning technology in the field of image analysis. Using live poultry excreta as an important basis for health judgement further improves the timeliness and accuracy of epidemic disease surveillance. A large number of live chicken excreta images were used for training and validation of the deep learning model. The results show that the system is able to successfully complete a series of processes such as automatic photography, analysis and identification with low errors. With the increase in the number of training samples in the subsequent image set, the system has the ability to further generalise and upgrade and expand, and the effect indicators are more satisfactory, which has good application prospects in the field of live poultry transport epidemic disease surveillance.