• 

  raw input

• 

  tested output

DotSpeak

Inspiration

Millions of visually impaired individuals rely on Braille for reading and communication. While Braille enables independence, many caregivers, teachers, volunteers, and family members cannot read it. We wanted to build a bridge between Braille users and non-Braille users through AI and computer vision.

DotSpeak was created to convert real physical Braille captured through a camera into readable English text and spoken output in real time, making information more accessible and inclusive.

What it does

DotSpeak captures images or video of physical embossed Braille using a webcam or mobile camera.

The system:

• Detects Braille dots from camera input
• Identifies Braille cell patterns
• Converts Braille patterns into English characters
• Generates readable text output
• Provides text-to-speech audio feedback
• Supports real-time or near real-time recognition

The goal is to help users quickly understand Braille content without requiring Braille literacy.

How we built it

The solution combines Computer Vision and Machine Learning techniques.

Workflow

1. Capture image/video from camera
2. Preprocess image using OpenCV
3. Detect Braille dot regions
4. Group dots into Braille cells
5. Recognize Braille patterns
6. Translate patterns into English text
7. Generate speech output using text-to-speech

Challenges we ran into

• Detecting small Braille dots under varying lighting conditions
• Handling camera blur and perspective distortion
• Correctly segmenting Braille cells
• Improving recognition accuracy for handwritten and embossed Braille
• Achieving real-time performance with limited computational resources

Accomplishments that we're proud of

• Built a working prototype for physical Braille recognition
• Successfully converted detected Braille into English text
• Added accessibility-focused speech output
• Developed a scalable pipeline that can be extended to multiple languages

What we learned

During development, we gained practical experience in:

• Computer Vision
• Object Detection
• Image Processing
• Accessibility Technology
• Machine Learning Model Integration
• Real-Time Application Development

What's next for DotSpeak

• Multi-language support
• Mobile application deployment
• Offline edge-device processing
• Improved recognition accuracy
• Smart camera guidance for visually impaired users
• Cloud-based accessibility services

DotSpeak demonstrates how AI can be used to make information more accessible and create meaningful social impact for visually impaired communities.

Built With

• accessibility
• computer
• github
• learning
• machine
• numpy
• opencv
• python
• text-to-speech
• vision
• yolo