A wearable assistive technology device that converts hand gestures into readable text in real time — bridging the communication gap for deaf and mute individuals.
The Smart Glove is a low-cost, portable, wearable system that detects finger bending patterns using flex sensors and translates them into alphabet letters displayed on a 16×2 LCD screen. Each unique combination of bent and straight fingers forms a 5-bit binary code that maps to a specific letter (A–Z), enabling real-time gesture-to-text communication.
This project was developed as part of the Community Engagement Project (SE, 1st Semester) at Dr. D. Y. Patil Institute of Technology, Pimpri, Pune — Department of Artificial Intelligence and Data Science (2025–2026).
| Name | Roll No |
|---|---|
| Ishita Bhoir | SAI&DA02 |
| Tanuja Naphade | SAI&DA05 |
| Vaibhavi Shinde | SAI&DA15 |
| Sameer Talekar | SAI&DA18 |
Project Guide: Mrs. Chetana Shravage
HOD AI & DS: Mrs. Shubhangi Vairagar
- Develop a wearable glove with flex sensors to detect finger bending and hand gestures
- Process analog sensor data using Arduino Uno with noise reduction via signal conditioning
- Map gesture combinations to alphabet characters using a binary-coded logic system
- Display detected letters on a 16×2 LCD in real time
- Build an affordable, portable communication aid for deaf and mute individuals
- Reduce dependency on interpreters by enabling direct gesture-based interaction
| Component | Specification |
|---|---|
| Microcontroller | Arduino Uno R3 |
| Flex Sensors | 5 units, 10kΩ base resistance |
| Display | 16×2 LCD (LiquidCrystal interface) |
| Resistors | 10kΩ (voltage divider per sensor) |
| Filtering Method | Median + EMA filtering |
| Power Supply | 5V DC via USB |
| Communication | Serial 9600 bps (testing) |
| Simulation Tool | Tinkercad |
| Programming Tool | Arduino IDE |
- Arduino IDE — firmware development
- Tinkercad — circuit simulation and validation before physical prototyping
- Serial Monitor — real-time debugging of sensor values
Five flex sensors (one per finger) detect bending. Each sensor acts as a variable resistor — resistance changes as the finger bends, producing a varying analog voltage read by the Arduino's ADC pins (A0–A4).
Raw sensor data passes through three stages:
-
Median Filtering — takes 3 consecutive samples and selects the median to remove random spikes:
$$x_M = M(x_1, x_2, x_3)$$ -
Exponential Moving Average (EMA) — smooths rapid fluctuations while maintaining responsiveness (α = 0.4):
$$EMA_{new} = \alpha \times V_{current} + (1 - \alpha) \times EMA_{previous}$$ -
Dynamic Self-Calibration — at startup, each sensor sets its own baseline. A dynamic threshold accounts for noise:
$$Margin = Fixed\ Margin + (Noise \times K)$$
Each finger is assigned a single bit:
| Finger | Label | Bit Position |
|---|---|---|
| Thumb | F1 | LSB (bit 0) |
| Index | F2 | bit 1 |
| Middle | F3 | bit 2 |
| Ring | F4 | bit 3 |
| Pinky | F5 | MSB (bit 4) |
A bent finger = 1, straight = 0. The 5-bit combination maps to a letter:
| Binary Code | Letter | Binary Code | Letter |
|---|---|---|---|
| 00001 | A | 10001 | Q |
| 00010 | B | 10010 | R |
| 00011 | C | 10011 | S |
| 00100 | D | 10100 | T |
| 00101 | E | 10101 | U |
| 00110 | F | 10110 | V |
| 00111 | G | 10111 | W |
| 01000 | H | 11000 | X |
| 01001 | I | 11001 | Y |
| 01010 | J | 11010 | Z |
| 01011 | K | ... | ... |
| 01100 | L | ||
| 01101 | M | ||
| 01110 | N | ||
| 01111 | O | ||
| 10000 | P |
- Row 1 of LCD — displays the currently detected letter
- Row 2 of LCD — displays the accumulated word/sentence (up to 32 characters)
- Real-time translation — gesture to letter with ~100ms refresh rate
- Dynamic self-calibration — adapts to different users, finger lengths, and sensor placements at every startup
- Noise-resistant — triple-stage signal conditioning (Median + EMA + Dynamic Threshold)
- Word construction — letters string together to form complete words
- Offline & portable — no internet, no camera, no controlled lighting required
- Deterministic logic — fully debuggable, no black-box ML model
- Low cost — built entirely with basic electronic components
- Simulation-first — validated on Tinkercad before physical prototyping
- Assistive communication for hearing and speech impaired individuals
- Educational tools in special schools
- Hospital and clinic interactions
- Public service counters for accessibility
- Future integration into IoT and smart home systems
smart-glove-gesture-to-text/
│
├── code/
│ └── smart_glove.ino # Arduino firmware
│
├── docs/
│ ├── project_report.pdf # Full project report
│ └── ieee_paper.pdf # IEEE conference paper
│
├── images/
│ ├── glove_prototype.jpg # Final glove prototype
│ ├── tinkercad_circuit.png # Circuit simulation
│ ├── block_diagram.png # System block diagram
│ ├── lcd_yes.jpg # LCD displaying YES
│ ├── lcd_help.jpg # LCD displaying HELP
│ └── breadboard_testing.jpg # Initial breadboard setup
│
└── README.md
- Clone the repository
git clone https://github.com/Sameer0726152/smart-glove-gesture-to-text.git-
Open the code
- Open
code/smart_glove.inoin Arduino IDE
- Open
-
Connect hardware
- Wire flex sensors to analog pins A0–A4 via 10kΩ voltage dividers
- Connect 16×2 LCD to digital pins 12, 11, 5, 4, 3, 2
- Power via USB
-
Upload
- Select board: Arduino Uno
- Select correct COM port
- Click Upload
-
Use
- Hold the glove still for 2 seconds during calibration
- Bend fingers to form gestures
- Read the detected letter on the LCD
- Only supports preprogrammed single letters (A–Z)
- Calibration required for different users
- Rapid gesture transitions may cause misreads
- Does not support dynamic or motion-based gestures
- Physical sensor contact must be secure for accurate readings
- Full word and phrase mapping beyond single letters
- Speech synthesis module for audio output
- Bluetooth / Wi-Fi integration to send text to smartphone
- Machine learning for adaptive thresholds and complex gestures
- More ergonomic glove design with flexible PCB
- IoT integration for controlling smart devices via gestures
This project is submitted for academic purposes at Dr. D. Y. Patil Institute of Technology, Pimpri, Pune. All rights reserved by the authors.
We sincerely thank Mrs. Chetana Shravage for her expert guidance, constant support, and encouragement throughout this project. We also thank Mrs. Shubhangi Vairagar (HOD, AI & DS) for his support.
Smart Glove — Making communication inclusive, one gesture at a time. 🤝