Electrovista | Computer Vision

Obstacle avoidance robot

This is a robot that can detect and avoid obstacles in its path using ultrasonic sensors and camera vision.

Features

Raspberry Pi Camera / USB Camera: Captures real-time video for obstacle detection
Raspberry Pi: Processes video feed and applies Canny edge detection
Motor Driver Module: Controls motor movements based on detected obstacles
DC Motors: Enables the robot to navigate around obstacles

How it Works

Canny Edge Detection: Identifies obstacle edges in the video feed
Region-Based Edge Analysis:
- More edges on the left? → Turn right
- More edges on the right? → Turn left
- Edges on both sides? → Stop
- No edges detected? → Move forward
Motor Control via Raspberry Pi GPIO: Adjusts movement based on obstacle detection

Objection Detection using Thermal Imaging Camera

Find out more about this simulation here

Components Comparison

Components	Simulation model	Real-world model
Microcontroller	Arduino Uno3	Raspberry Pi4
Temperature sensor	TMP36 (simulated)	MLX90640 (thermal imaging sensor)
Display	16x2 LCD (non-I2C)	Laptop display
Cooling system	Not required	Heat sink and fan for Raspberry Pi
Power supply	Simulated 5V from Arduino	5V USB-C adapter for Raspberry Pi

Simulation Components

Arduino Uno: Acts as the microcontroller to process data and control the system
TMP36 Temperature Sensor: Simulates temperature readings and outputs an analog voltage
16x2 LCD Display: Displays the temperature and object classification
Potentiometer: Adjusts the contrast of the LCD display

Simulation Flow

Temperature Sensing:
- TMP36 sensor simulates temperature readings
- Outputs an analog voltage proportional to the temperature
- Arduino reads the voltage and converts it into Celsius
Object Classification:
- Temperature above 30°C → Classified as Human/Dog
- Temperature below 30°C → Classified as Car/Bike
Display Output:
- First Line: Displays object classification (e.g., "Human/Dog" or "Car/Bike")
- Second Line: Displays temperature value in Celsius (e.g., "Temp: 35.00 C")
Adjusting Contrast:
- Potentiometer adjusts LCD contrast
- Turning the knob makes the text clearer or dimmer

Tinkercad Interaction

Manually adjust TMP36 temperature values
Arduino processes new values and updates the LCD in real-time
Observe object classification changes based on temperature

Gesture controlled device switch

Find out more about this simulation here

A user wants to control a bulb in their room using hand gestures. Instead of pressing a traditional switch, they can use a simple hand wave or other gestures to turn the bulb on or off. This makes it easier to control the appliance without physically interacting with a switch, which is especially useful for people with limited mobility or for anyone who wants a more modern way to interact with their home appliances.

Raspberry Pi Setup

Raspberry Pi: Any model with camera support (e.g., Raspberry Pi 3 or 4)
Relay Module: Controls the table lamp based on detected gestures
Camera Module: Captures hand gestures for processing (via OpenCV or APDS-9960)
LCD Screen (Optional): Displays lamp status and other information
Power Supply: Powers Raspberry Pi and connected components
Wires & Connectors: Connects the relay and other components
Table Lamp: Connected to the relay module for gesture-based control

How It Works

Gesture Detection:
- Camera module captures real-time images of the user's hand
- OpenCV processes gestures like hand waves or specific hand shapes
- Example: Open hand → Turns light on, Fist → Turns light off
Control Logic:
- Raspberry Pi processes the camera feed to recognize gestures
- When a gesture is detected, Raspberry Pi triggers the relay module
- Relay is connected to GPIO pins, acting as a switch for the lamp
- Python script configures GPIO pins to send HIGH or LOW signals
Status Display:
- LCD screen (if connected) displays "On" or "Off" based on gesture recognition
Power Supply:
- Raspberry Pi requires a 5V power supply
- Relay module may need a separate power source depending on lamp voltage

Camera Setup and Gesture Recognition

Raspberry Pi camera module or external camera captures real-time images
Python script in OpenCV processes camera feed and detects gestures
Defines a region of interest (ROI) where hand detection occurs

Relay Control via GPIO

GPIO pins control physical hardware based on gesture detection
HIGH signal → Completes the relay circuit, turning lamp on
LOW signal → Opens the circuit, turning lamp off

COMPUTER

VISION

Project Head

Our Projects

Obstacle avoidance robot

Objection Detection using Thermal Imaging Camera

Gesture controlled device switch

Face Detection based door lock

Thinking of a cool project? We’d love to hear from you—get in touch!