CNN: Low-Cost High-Accuracy pH Detection using Machine Learning

Project Scope

This project focuses on creating a cost-effective and accessible pH level detection system by integrating machine learning with mobile technology. Leveraging a Convolutional Neural Network (CNN), the project enables precise pH classification using images of litmus paper and deploys the trained model to a custom-developed Android mobile application.

With a streamlined workflow encompassing data preparation, model training, and mobile deployment, the project provides an innovative alternative to traditional pH meters—eliminating the need for expensive equipment and promoting usability in resource-limited settings. Future developments aim to expand capabilities and enhance versatility for broader applications in water quality monitoring and beyond.

Highlights

• Low-cost pH level detection using machine learning

• Integration of TensorFlow Lite for mobile app deployment

• High-accuracy classification with Grad-CAM visualizations

• Mobile app "The App(H)" for real-time field measurements

• Applications in water quality monitoring and environmental science