International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 08 | Aug 2025
p-ISSN: 2395-0072
www.irjet.net
Automated Skin Disease Diagnosis Using DINO Vision Transformers and OpenCV Samiksha Yadav1, Koppada Prudhvi Vinayak2 1B.Tech Student, Department of Computer Science and Engineering (AI & ML), Pragati Engineering College, East
Godavari, India
21B.Tech Student, Department of Computer Science and Engineering (AI & ML), Pragati Engineering College, East
Godavari, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - This research provides an efficient, cost-effective,
detection of skin conditions, particularly in areas with limited medical resources.
and affordable solution for automated skin disease diagnosis through state-of-the-art image processing and deep learning methods. Our objective is to overcome costly and scarce dermatologic diagnostic impediments, especially for underserved communities. We implemented a web-based application that utilizes OpenCV to extract important features from skin images—involving color, texture, and shapes—and applies a self-supervised Vision Transformer (ViT) model trained through DINO for robust and precise classification. The system provides an easy-to-use interface for uploading skin images, whereby images are automatically processed and analyzed in real time. Utilizing a publicly accessible, annotated dermatoscopic image dataset, our model was about 92% accurate in predicting multiple skin ailments. The prime contribution of this research lies in combining powerful computer vision algorithms with an easily scalable and easyto-use web application, which forms an efficacious tool for early diagnosis and screening of skin diseases that can be easily implemented in real-world clinical and remote healthcare environments.
1.1 Objectives The objective of this project is to design and develop a webbased application for automated skin disease diagnosis that is both accessible and user-friendly. The system leverages advanced image processing techniques using OpenCV to extract relevant features from skin images, including color, texture, and shape. For accurate and robust classification of skin diseases, a deep learning model based on the Vision Transformer (ViT) architecture is implemented, pretrained using the DINO self-supervised paradigm. The performance of the system is rigorously evaluated on publicly available, annotated dermatoscopic image datasets using standard metrics such as accuracy, precision, recall, and F1-score. Ultimately, this solution aims to be scalable and cost-effective, enabling its use in real-world clinical and remote healthcare environments for early screening and diagnosis of skin conditions.
1.2 Literature Survey
Key Words: Skin disease diagnosis, computer vision, OpenCV, Vision Transformer (ViT), DINO, deep learning, web application, medical image analysis, automated diagnosis, healthcare accessibility
With an ever-growing incidence of skin diseases around the world, considerable research interest lies in automated dermatologic diagnosis. Current methods are mostly based on handcrafted feature extraction and traditional machine learning algorithms, yet these are frequently hindered by subjective human interpretation and issues of scalability.
1.INTRODUCTION Skin conditions affect millions worldwide, with over 900 million individuals experiencing some form of skin disorder at least once in their lifetime. Timely and precise diagnosis is essential for effective treatment, but conventional methods often demand specialized knowledge and expensive equipment, limiting accessibility in many regions.
With recent advancements in deep learning, particularly in convolutional neural networks (CNNs), remarkable performance in medical image analysis, such as skin lesion classification, has emerged. CNN-based networks are capable of learning automatic discriminative features from raw pixel information without handcrafted features, which promotes better
Recent advancements in computer vision and artificial intelligence offer scalable and objective diagnostic solutions. Utilizing OpenCV for image feature extraction and a Vision Transformer (ViT) model trained with the DINO selfsupervised learning approach, we propose an automated system for classifying skin diseases. This platform facilitates easy user registration, secure image uploads, and fast diagnosis, providing an affordable and efficient tool for early
© 2025, IRJET
|
Impact Factor value: 8.315
ViT architectures have become strong competitors to CNNs for classification on images in more recent times. ViTs employ self-attention for extracting high-order patterns and long-range context in images, achieving state-of-the-art performance across several benchmarks. Self-supervised
|
ISO 9001:2008 Certified Journal
|
Page 377