Vision-I is a vision enhancement module specifically designed for blind individuals. It allows blind users to utilize an application that sends real-time frames to a laptop-based wireless networked system. The system employs real-time object detection using the SSD_MOBILENET algorithm and TensorFlow APIs. Its core feature includes approximate distance calculation and voice-based wireless feedback generation based on the distance calculation. Vision-I aims to make the lives of blind individuals easier, more efficient, and reliable by providing wireless voice-based feedback regarding the proximity of objects, thereby enhancing their navigation experience. Additionally, the same system can be used for obstacle detection.
- Easy integration
- Voice activation
- Real-time interface
- Compatibility with Android mobile webcams
To set up Vision-I, follow these steps:
- Fork the repository to your profile.
- Clone the repository to your local machine using Git.
- Install dependencies by running
pip install -r requirements.txt. - Follow the YouTube tutorial provided for further guidance.
- Run the [MainFile] webcam_blind_voice.py to start using the system.
There are some improvements and additional features planned for Vision-I:
- Make it dynamic over cloud systems for enhanced scalability and accessibility.
- Provide features to input the class of object for more precise feedback.
Feel free to contribute to Vision-I and help make it even more accessible and user-friendly for blind individuals.