Practical Mobile Based Services for Identification of Chicken Diseases from Fecal Images
Issued Date
2024-01-01
Resource Type
ISSN
21593442
eISSN
21593450
Scopus ID
2-s2.0-105000415320
Journal Title
IEEE Region 10 Annual International Conference, Proceedings/TENCON
Start Page
108
End Page
111
Rights Holder(s)
SCOPUS
Bibliographic Citation
IEEE Region 10 Annual International Conference, Proceedings/TENCON (2024) , 108-111
Suggested Citation
Silapachote P., Srisuphab A., Damkham W., Korkiattrakool P., Songdechakaivut K. Practical Mobile Based Services for Identification of Chicken Diseases from Fecal Images. IEEE Region 10 Annual International Conference, Proceedings/TENCON (2024) , 108-111. 111. doi:10.1109/TENCON61640.2024.10902790 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/108622
Title
Practical Mobile Based Services for Identification of Chicken Diseases from Fecal Images
Author's Affiliation
Corresponding Author(s)
Other Contributor(s)
Abstract
Poultry farming is a vital component in a food chain. Health of chickens in the farms directly plays an important role in both quality and safety of all chicken products. Accurate diagnosis of poultry diseases from chicken feces using standardized polymerase chain reaction is relatively very expensive, making it rather challenging especially for small local farms. To tackle this, we developed a practical mobile-based service that aimed to provide local farmers an easy-to-use tool capable of preliminary identification of common diseases chickens may have from their fecal images. Our system was deployed as an official Line account on a familiar Line application that most farmers have on their mobile phones. Trained and evaluated on a large open database of chicken fecal images, it achieved a segmentation mean average precision of 86.49 % and classification accuracy of 95.93 %. Ours also, with high confidence, correctly identify healthy images taken by local farmers.