Medication mismanagement is a major cause of health complications, especially among elderly and visually impaired patients. Traditional medicine cabinets rely entirely on manual identification, which can lead to wrong medicine intake.
The Smart Medicine Cabinet solves this problem by :
- Automatically identifying medicine type using computer vision
- Providing feedback through display or audio
- Improving medication safety and organization
The system is intentionally limited to detecting:
- Bacillus Clausii Spores Suspension
- Insulin (vial or pen)
- Eye Drops
( This controlled scope improves accuracy and reliability. )
System Components
Hardware Components :
- ESP32 / ESP32-CAM
- Camera Module
- Medicine Cabinet (Compartments)
- Speaker / Buzzer (Audio Feedback)
- OLED / LCD Display (Optional)
- LED Indicators
- Power Supply (5V / Battery)
Software Components :
- Python
- OpenCV
- Akaze Algorithm / YOLO (trained model)
- ESP32 Firmware
- Serial / Wi-Fi Communication
CIRCUIT DIAGRAM
Applications :
- Smart homes for elderly care
- Hospitals and clinics
- Assisted living centers
- Medicine safety systems
- Healthcare automation projects
Medicine Categories Detected :
Category | Description |
Bacillus Clausii Spores Suspension | Clear liquid vial with white cap |
Insulin | Insulin vial or insulin pen |
Eye Drops | Small dropper bottles |
( Other medicines are intentionally ignored to avoid false detection. )
Advantages :
- Reduces medication errors
- Simple and user-friendly
- Limited detection improves accuracy
- Cost-effective system
- Supports visually impaired users
Connection Description (Wiring Map) :
Component | Connection | Description |
Camera | ESP32-CAM | Captures medicine image |
Speaker | GPIO Pin | Audio output |
Display | I2C Pins | Shows detected medicine |
LED | GPIO | Status indication |
Power | 5V / GND | System power |
Wi-Fi | ESP32 | Communication with the Python system |
Working Principle :
- Image Capture
- When a medicine is placed in front of the camera, an image is captured.
- Computer Vision Processing
- Image is processed using OpenCV.
- Trained model (YOLO / CNN) classifies the object as:
- Bacillus Clausii Spores Suspension
- Insulin
- Eye Drop
- Decision Logic
- If medicine belongs to the allowed categories → accepted
- If unknown → ignored or a warning given
- User Feedback
- Audio: “Insulin detected”, etc.
- Display: Medicine type shown on screen
- LED indication for confirmation
- Optional Storage Action
Cabinet opens the correct compartment (future enhancement).
Testing the Hardware and Software
Camera Test :
- Verify image clarity and focus.
- Test under different lighting conditions.
Model Test
- Test with sample medicines.
- Check accuracy for each category.
Communication Test
- Verify data transfer between the Python system and the ESP32.
Full System Test
- Place each medicine type and confirm correct identification.
( Test rejection of unknown objects. )
Troubleshooting :
Issue | Cause | Solution |
Wrong detection | Poor lighting | Improve lighting |
Camera not working | Power issue | Check voltage |
Audio not playing | Speaker wiring | Check the GPIO pin |
False positives | Model confusion | Retrain with a better dataset |
ESP32 resets | Power drop | Use a stable supply |
( The Smart Medicine Cabinet with Computer Vision is an intelligent healthcare system designed to automatically identify and manage medicines using a camera and image processing techniques. The system is trained to detect only three medicines – insulin vials, bacillus clausii spores suspension, and eye drops. Using computer vision and a microcontroller-based system, the cabinet assists users by identifying medicines, giving audio or display feedback, and reducing medication errors. This project is especially useful for elderly people, patients, and healthcare facilities. )