Video Call Intercom with Vibration Sensor for Deaf Communication

• As a deaf user, I can receive vibration alerts and see incoming calls on screen.

• As a family member or visitor, I can initiate a video call and communicate in real time.

• Acceptance: Vibration motor activates within one second of a call; video call connects within three seconds on a broadband network.
  

Module 1: Server Management

FastAPI server tracks connected devices and their IP addresses.

Module 2: Video Communication

Python sockets stream video frames and data packets for low-latency communication.

Module 3: Gesture Interpretation

Mediapipe monitors hand movements and matches them with a trained dataset to recognize 24 common words and phrases.

Module 4: Call Alerts

Vibration motor activates whenever an incoming call is detected.

Module 5: User Interface

Tkinter-based touch interface shows connected devices and lets users place or accept calls.

• Hardware: Raspberry Pi 4, camera module, vibration motor, XPT2046 touchscreen

• Backend: FastAPI

• Frontend/UI: Python Tkinter

• Libraries: OpenCV, Mediapipe, GPIO, Socket

Option B (Web UI Alternative):

• Backend: Django/Channels for signaling

• Frontend: React or Vue for browser-based dashboard

• Optional: WebRTC for browser-based video streaming
  

• Call connection latency

• Gesture recognition accuracy (%)

• Uptime of intercom service

• Number of successful vibration alerts per day
  

• Week 1: Finalize SRS, wiring schematic, and database (if needed)

• Week 2: Set up FastAPI server and Pi camera streaming

• Week 3: Integrate vibration motor and develop basic UI

• Week 4: Add sign-language recognition and speech-to-text

• Week 5: Test and tune performance

• Week 6: Documentation, final testing, and presentation
  

• Students/Capstone Teams seeking a socially impactful IoT project

• Instructors wanting a ready-to-deploy pilot kit

• Institutions looking for a scalable, low-cost communication aid