INR_Classifier

Microsoft-Future-Ready-Talent-Virtual-Internship-Project

Project Title

Automated Indian Rupee (INR) Denomination Classifier using Azure Custom Vision Service..

Industry: Finance Training Images link:https://drive.google.com/drive/folders/1e-DNtL04nl3Il1uW-gca1nDloD4ClCQH?usp=sharing

Core Azure Services

Azure App Service: Used for hosting the web application and API endpoints. Visual Studio Code: Integrated development environment (IDE) for coding, testing, and deploying the project.

Azure AI Service

Azure AI Custom Vision Service: Leveraged for training and deploying the custom image classification model for INR denomination detection.

Problem Statement

The manual process of identifying and classifying Indian Rupee (INR) denominations in large volumes can be time-consuming and error-prone. This project aims to automate the classification of INR currency notes using machine learning, providing an efficient and accurate solution for financial institutions and businesses.

Key Features

Automated INR Denomination Classification User-Friendly Web Interface Real-time Prediction Two Image Upload Methods: Capture Image using OpenCV (cv2) Upload Image (JPEG, PNG, etc.) Azure App Service Hosting Integration with Visual Studio Code for development and deployment

Project Workflow

Set Up Environment 1.Clone the GitHub repository: git clone https://github.com/Aasthajain02/INR_Classifier.git

2.Navigate to the project directory: cd INR_Classifier

3.Install dependencies: pip install -r requirements.txt

4. Set up environment variables in config.py for Azure Custom Vision Service credentials.

Training and Deployment Use Azure Custom Vision Service to create and train a custom image classification model for INR denominations. Retrieve the prediction endpoint, prediction key, project ID, and model name from Azure Custom Vision Service. Update config.py with the obtained credentials.

Screenshots

1. Azure AI Custom Vision Service Training Dataset Description: 50 images of each currency 10,20,50,100,200,500,2000 are uploaded

3. Accuracy of Quick training

4. Prediction image

   

Conclusion

The INR_Classifier project demonstrates an end-to-end workflow for building a web-based INR denomination classifier using Azure services. By leveraging Azure Custom Vision Service, the project provides an efficient and scalable solution for automating the classification of INR currency notes.