Salmonella risk prediction in poultry farms via deep learning image classification models, cross-region validation, and edge computing

Authors: BODEMPUDI, VENKAT - University Of Georgia; LI, GUOMING - University Of Georgia; Oladeinde, Adelumola; Rothrock Jr, Michael; AGGREY, SAMUEL - University Of Georgia; LIU, TONGSHUAI - University Of Georgia; KOTA, SAI - University Of Georgia; YUAN, GENG - University Of Georgia; NALLA, SRAVAN SAI RAHU - University Of Houston

Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/23/2026
Publication Date: 11/15/2026
Citation: Bodempudi, V., Li, G., Oladeinde, A.A., Rothrock Jr, M.J., Aggrey, S., Liu, T., Kota, S., Yuan, G., Nalla, S.L. 2026. Salmonella risk prediction in poultry farms via deep learning image classification models, cross-region validation, and edge computing. Computers and Electronics in Agriculture. https://doi.org/10.1016/j.compag.2026.112116.
DOI: https://doi.org/10.1016/j.compag.2026.112116

Interpretive Summary: Poultry products are recognized as one of the primary sources of salmonellosis. Rapid on-farm prediction tools for Salmonella will support timely and precise control of the infection at pre-harvest and thus improve food safety. This study introduces a system for Salmonella prediction using chicken fecal images. The system leverages a sophisticated machine learning model for Salmonella infection classification. Following a thorough optimization, the model developed exhibited a modest predictive accuracy of 85% and was made available in the Android Studio and Apple iOS environments. Compared to other currently available poultry disease decision support systems, this tool is simpler, safer, and uses fewer computational resources. Furthermore, it allows users to custom train the system, using their own data to improve the classification of unseen fecal images. Therefore, it is a promising tool that farmers and veterinarians in North America can use and optimize to make informed decisions about the Salmonella status of their flock.

Technical Abstract: Salmonella remains a major cause of foodborne illnesses that are commonly associated with the consumption of contaminated poultry products. Therefore, there is a strong emphasis and a crucial need to reduce Salmonella across the farm to fork continuum. However, there are no easy and rapid on-farm prediction tools to estimate if a flock carries Salmonella. In this study, we used deep learning to develop a Salmonella risk classification model for chicken feces. The model was developed using a two-part framework combining deep learning and tree-based classification techniques. In the first stage, deep learning models such as ResNet, Xception, VGG, and others were employed to extract high-level features from chicken fecal images. In the second stage, these features were used by decision tree classifiers including Random Forest and Gradient Boosting for binary classification into Salmonella PCR Positive or Negative. The model was trained and cross-validated on two datasets of an American dataset with 1149 images from a controlled multi-serovar Salmonella challenge study and African dataset with 4025 publicly available images. This integrated approach achieved up to 85% classification accuracy when validated across geographical regions. The developed framework was integrated into a user-friendly mobile and web application and allows users to custom train the deep learning classifiers to improve the classification of unseen fecal images. Although the model was primarily validated with the same dataset, it can still be useful as a decision support tool for Salmonella control and interventions, thereby enhancing food safety.