Explore both iterations in action:
▶️ ESP32 + OpenCV-based Posture Corrector (2nd Iteration)▶️ Arduino-based Posture Corrector (1st Iteration)
📁 Watch the Demo Videos on Google Drive
A wearable device that uses real-time sensor feedback and computer vision to promote proper posture.
This device monitors upper back and neck posture using an MPU-6050 sensor and a custom OpenCV model. Upon detecting poor posture, it emits an audible alert via a buzzer to notify the user.
- Built using an Arduino Uno R3, MPU-6050, and a breadboard.
- Detected angular deviations in back posture and triggered a buzzer during slouching.
- Upgraded to an ESP32 microcontroller for wireless communication and improved performance.
- Integrated an OpenCV + MediaPipe model to track neck posture using webcam input and calculate head tilt angles.
- Real-time communication between OpenCV and ESP32 triggers the buzzer when poor neck posture is detected.
- Designed a 3D-printed wearable enclosure using SolidWorks to house the electronics.
- 📂 Computer vision source code: OpenCV Posture Model
- Designed a custom PCB to replace the breadboard, enhancing compactness, reliability, and ease of assembly.
- A new SolidWorks 3D enclosure is in development for the updated hardware version.