Published April 20, 2025 | Version v1

CANINE INJURY DETECTION: A COMPARATIVE ANALYSIS OF IMAGE CLASSIFICATION MODELS

Description

The swift and accurate detection of canine injuries 
is crucial for timely medical intervention and enhancing public 
safety. This study presents a comprehensive analysis of three 
image classification models—Convolutional Neural Networks 
(CNN), Random Forest, and Support Vector Machines (SVM)—
 for detecting injuries in dogs. Utilizing a meticulously curated 
dataset of over 2000 images, the research emphasizes robust data 
gathering, cleaning, and preprocessing techniques. Experimental 
results reveal that CNN outperforms with an accuracy of 75.78%, 
followed closely by Random Forest and SVM. The findings 
highlight the strengths and limitations of each model, offering 
valuable insights into their applicability in real-world scenarios. 
This research underscores the potential of advanced machine 
learning models in promoting animal welfare and lays a 
foundation for further advancements in the field of computer 
vision applications in veterinary care. 

Files

CANINE INJURY DETECTION.pdf

Files (409.7 kB)

Name Size
md5:8c438e2794c9fac605aef3acf44036b4
409.7 kB Preview Download