B.E.R.T.H.A.

Inspiration

The inspiration for B.E.R.T.H.A. stemmed from a personal connection. Nick's sister was diagnosed with a brain tumor that compromised her vision, making her legally blind. This fueled our passion for creating a tool to aid those with visual impairments in their daily lives.

What it does

B.E.R.T.H.A. assists visually impaired individuals by using object detection to highlight everyday objects and potential hazards. It processes real-time video footage via a Raspberry Pi 4B and communicates the information back to the user through Bluetooth audio.

How we built it

We built B.E.R.T.H.A. by combining YOLO's object detection library with a Raspberry Pi 4B, equipped with a 1080p webcam. The device captures and processes video footage, then transmits the information to the user via Bluetooth audio, all housed in a 3D-printed casing.

Challenges we ran into

We faced challenges with the initial speed and accuracy of the system, as YOLO’s v3 model caused significant lag in real-time object detection. Switching to YOLOv5 improved both speed and accuracy, allowing for practical real-time detections.

Accomplishments that we're proud of

We are proud of successfully integrating a robust object detection system that operates in real-time and provides accurate feedback to users. Additionally, creating a functional and portable design that can genuinely aid visually impaired individuals is a significant accomplishment.

What we learned

We learned the importance of optimizing both hardware and software to achieve real-time performance. Additionally, we gained experience in using advanced object detection libraries and integrating various sensors to enhance the functionality of our project.

What's next for B.E.R.T.H.A.

Next, we plan to develop a more sleek and compatible casing, upgrade the camera for higher video quality, and implement custom-trained YOLO models for specialized detection. We also aim to integrate new sensors and GPS compatibility to enhance the device's capabilities.