+27 60 810 9826 leokuyper@gmail.com
Select Page

Project

Smart Glasses

The SmartGlasses project is an AI-based initiative designed to assist individuals with visual impairments. The system is built using Python3 and OpenCV and operates on a Raspberry Pi 4 Model B. The core functionalities of the system include aiding visually impaired individuals, assisting people with Prosopagnosia (Face Blindness), face detection with attributes, scene detection, and OCR (Text recognition). Additional features include voice recognition and voice feedback. The project was initially intended to run all recognitions simultaneously, but due to performance issues, the functionalities were split into different modes, allowing the user to switch between them. The project also includes a multithreaded approach to how OpenCV works and processes information, allowing for efficient use of the Raspberry Pi’s cores

Goal

Smart Glasses

The system, running on a Raspberry Pi 4 Model B, uses OpenCV for image processing and Google’s voice services for voice recognition and feedback. It offers functionalities like face detection, scene detection, and OCR (Text recognition). It used a angled mirrors to allow your eyes to focus on the display to give you live data and feedback.

Enhance Accessibility

Improve the system’s ability to assist visually impaired individuals and those with Prosopagnosia (Face Blindness) by refining the voice feedback and recognition features. This could involve enhancing the accuracy of voice recognition and expanding the range of commands the system can understand and respond to.

Expand Functionality

Develop and integrate additional features into the system. One potential future functionality mentioned in the documentation is face recognition, which would allow the system to identify specific people. This could greatly enhance the user experience, particularly for individuals with Prosopagnosia.

Optimize Performance

Continue to optimize the system’s performance on the Raspberry Pi 4 Model B. This could involve further refining the multithreaded approach to how OpenCV works and processes information, or exploring other methods of improving the system’s efficiency and responsiveness.