This project aims to create a wearable object detection system using machine learning and computer vision. It will leverage the Coral Dev Board’s TPU for efficient inference, coupled with OpenCV to classify objects such as "Cross" vs. "Don’t Cross" for real-time decision-making.
Arjun Rao & Armaan Gera
- TensorFlow - A machine learning library that is used to train and run models. We use this library to export our model to a .tflite file and run it on the Google Coral Dev Board.
- OpenCV - A computer vision library that is used to process images and video.
- Flutter - A UI toolkit that is used to create the mobile app that interacts with our wearable solution via Wifi/Bluetooth.
- Google Coral Dev Board - A computer with a Mendel Linux OS that is used to run our machine learning model.
- Arducam 5MP Wide Angle Camera - The camera that is used in our wearable solution.
- Screen Reader Compatible (Web Content Accessibility Guidelines)
- Wearable solution that interacts with an app through a Wifi signal or Bluetooth
- Camera integrated with some form of wearable hardware (glasses)
- Real-time object detection or classification for improving the individual's spatial awareness
- Color-neutral UI with minimal number of pages (easy and intuitive navigation)
- Developed for iOS and Android devices.
| Status | Date | Description |
|---|---|---|
| Started | 9/9/2024 | Description of Current Status |
| Completed | 9/17/2024 | Description of Completed Feature |
| Testing | 9/17/2024 | Unit Tests Needed |
| Bugged | 9/17/2024 | Link to Git Issue |
