A comparative study of deep learning models for automated detection of Lumpy Skin Disease (LSD) in Cattle

Abdul Raziff, Abdul Rafiez and Mohd Pozi, Muhammad Syafiq and Mohamed, Raihani and Samsudin, Anjas Asmara (2026) A comparative study of deep learning models for automated detection of Lumpy Skin Disease (LSD) in Cattle. In: 2025 10th International Conference on Information and Communication Technology for the Muslim World (ICT4M), 26-27 November 2025, KUALA LUMPUR, Malaysia.

	PDF - Published Version Restricted to Repository staff only Download (947kB) | Request a copy
Preview	PDF - Supplemental Material Download (141kB) | Preview

Abstract

Lumpy Skin Disease (LSD) is a highly contagious viral illness affecting cattle, leading to significant economic losses in the livestock industry due to reduced milk yield, weight loss, infertility, and mortality. Early and accurate diagnosis is crucial for containment and treatment. Traditional diagnostic methods are time-consuming and require expert veterinarians, which may not be accessible in rural areas. This study proposes and evaluates a range of deep learning models for automated identification of LSD from cow images. We compare the performance of six deep learning architectures: CNN-TensorFlow, CNN-SqueezeNet, CNN-Inception, CNN-Xception, MobileNet, and PyTorch based CNN, alongside traditional machine learning classifiers such as SVM, MLP, Logistic Regression, J48, and an ensemble model. Using a dataset of 1,024 labeled images (split 80:20 for training and testing), our experiments show that Xception achieves the highest accuracy of 90.15%, followed by MobileNet (89.16%) and Inception (88.18%). The original CNN-TensorFlow model achieved 84.73%, outperforming SVM (81.28%) and MLP (67.49%). Notably, SqueezeNet failed to de- tect any positive cases (recall = 0), highlighting model sensitivity to architecture choice. The results demonstrate that deep learning, particularly transferlearning-capable models like Xception and lightweight MobileNet, offers a robust solution for automated LSD detection. This work lays the foundation for developing real- time, mobile-based diagnostic tools for farmers and veterinarians, contributing to sustainable livestock management.

Item Type:	Proceeding Paper (Other)
Uncontrolled Keywords:	lumpy skin disease, deep learning, convolutional neural networks, machine learning
Subjects:	K Law > KBP Islamic Law > KBP1 Islamic law.Shariah.Fiqh > KBP490 Furūʻ al-fiqh. Substantive law. Branches of law. > KBP3123 Animal protection. Animal welfare. Animal rights T Technology > T Technology (General) > T55.4 Industrial engineering.Management engineering. > T58.5 Information technology T Technology > TS Manufactures > TS1950 Animal products
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Information and Communication Technology Kulliyyah of Information and Communication Technology Kulliyyah of Information and Communication Technology > Department of Information System Kulliyyah of Information and Communication Technology > Department of Information System
Depositing User:	Dr Abdul Rafiez Abdul Raziff
Date Deposited:	14 May 2026 12:26
Last Modified:	14 May 2026 12:26
Queue Number:	2026-05-Q3319
URI:	http://irep.iium.edu.my/id/eprint/128949

Actions (login required)

View Item