A Flutter Mobile Application to classify the 72 Dogs Breed(including Himalayan Sheepdog i.e., Bhote Kukur).
Dog breed classification is a challenging task due to the wide variety of dog breeds with subtle differences in appearance. Convolutional Neural Networks (CNNs) have shown remarkable performance in image classification tasks, making them an ideal choice for dog breed recognition. In this project, we propose a CNN-based approach for dog breed classification.
The project 'Dog Breed Classification Using CNN' aims to develop a mobile application using the Flutter framework for accurate identification of dog breeds from images. This report presents the methodology, implementation details, and evaluation of the developed application. The Convolutional Neural Network (CNN) model was trained using a comprehensive dataset of dog images to enable accurate breed classification. The application provides an intuitive user interface, allowing users to capture or upload dog images for real-time breed identification. Evaluation results demonstrate the effectiveness of the CNN model, achieving high accuracy in breed classification. Additionally, the report discusses the challenges faced during the development process and proposes potential areas for future enhancements. The 'Dog Breed Classification Using CNN' project showcases the integration of deep learning techniques, mobile app development, and breed recognition, offering practical applications in various domains such as pet care, animal welfare, and canine research.
The American Kennel Club (AKC) recognizes around 197 dog breeds, each with unique physical characteristics and temperament. Accurately identifying the breed of a dog can be challenging and there is a need for a computer vision system that can classify dogs by their breed. Traditional image recognition techniques, such as feature engineering and Support Vector Machines, have been used to classify dog breeds with some success, but they often require manual feature selection and are not scalable for large datasets. CNNs have shown great promise in solving image classification problems, including dog breed classification, without requiring manual feature selection.
• Breed Classification
• Real-Time Classification
• Offline Functionality
The screenshot of output through the gallery and live feeds are given as below. The Application predicts the breed along with the accuracy level in percentage as shown below.
