Cattle Breed Classification Using Superpixel-Driven Graph Neural Network

View at publisher Summarise with VIRAT
Authors: Devansh Bhatnagar, Saurabh Natha Gharat, Manas Choudhari, Rekha Kaushik
Cited by: 0
References: 12

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Correctly identifying breeds of cattle in India is critical to supporting improved livestock management practices, facilitating selective breeding practices, and optimizing dairy production efficiency. The aim of this article is to describe the development of an automated framework utilized for breed identification. This was based on the application of image processing and deep learning models—the framework made use of preprocessing of the cattle images with segmenting techniques labelled as Simple Linear Iterative Clustering (SLIC) and color enhancement algorithms in order to improve the quality of the cattle images while preserving the natural visibility of the breed—as shown in the original cattle images. The features extracted using the processing techniques were then provided as input to a ResNet-based feature extractor to form deep visual representations from the images. The visual representations were provided to a Graph Neural Network (GNN) that was able to model the complex spatial relationships of segmented regions of the cattle to improve recognition accuracy. The model was trained and tested on labeled cattle breed images for a total of 30 epochs with a consistent accuracy across the epochs and fast convergence. This research has demonstrated the successful integration of SLIC, ResNet, and GNN components into an automated breed classification framework that can provide a reliable and scalable automated breed identification component of smart livestock management systems.

Fields of Study & Subjects

Smart Agriculture and AI Advanced Neural Network Applications Livestock Farming and Management

Publication details

Journal / source
IEEE International Students' Conference on Electrical, Electronics and Computer Science (SCEECS)
Pages
1–5
Publication date
31 January 2026
Language
EN