Medicine OCR: Pharmacy AI Scanner
Tailored for a SAAS client with over 300 pharmacy retailers, this solution revolutionizes the way pharmacies manage their inventory and orders. Using advanced AI and smartphone technology, we developed a system that seamlessly scans and updates medicine details, significantly enhancing operational efficiency. This project shows our commitment to creating cutting-edge, custom tech solutions that integrate smoothly with existing systems.
The Challenge: Speed & Accuracy
The primary challenge was to develop a system that could accurately and efficiently scan medicine details under various conditions such as differing smartphone performances, lighting variations, and movements. This needed to be achieved while ensuring seamless integration with the client's existing software and maintaining high accuracy in data capture.
- Diverse Medicine Packages & Labels
- Varying smartphone performance and specifications
- Varying lighting conditions and movements
- High accuracy requirement in data detection
- Integration with existing client software
Innovative AI-Powered Solution
Our approach involved a two-fold strategy: developing an intuitive Android app and refining a machine learning model. The app featured a camera preview for automatic detection, scanning, and uploading of medicine details. On the backend, we experimented with various models, ultimately training a machine learning model on a dataset of 1000 images. This model was uniquely designed to recognize key details such as expiry dates and batch codes with 99.5% accuracy.
- Intuitive app development for Android with automatic detection
- Extensive testing with various machine learning models
- Customized training of the ML model for high-accuracy OCR and NER
- Seamless integration with client’s existing software using APIs
The Impact
This solution transformed pharmacy operations, enhanced data accuracy and operational efficiency, setting a new standard in the pharmaceutical retail industry.
- 99.5% accuracy
- 80% time-saving for the end users
- Intuitive User Interface built
Project Information
Completed in
6 Weeks
Category
Pharmacy, Computer Vision
Tech Stack
Kotlin, Python, OCR, ML Kit, OpenCV
