Intermediate Level • Lesson 8

📷 Camera and Vision Basics

⏱️ 50 minutes 📚 Intermediate 🎯 Computer Vision

📷

Camera and Vision Basics

Learn how to use the Raspberry Pi Camera to take pictures and detect motion - this is the first step toward human tracking!

Raspberry Pi Camera

The Raspberry Pi Camera is a small camera module that connects directly to your Raspberry Pi. It can take photos, record videos, and with the right software, detect objects and people!

💡 Real-World Example: Security cameras use computer vision to detect people, cars, and movement. Your Raspberry Pi can do similar things!

Setting Up the Camera

Hardware Setup:

Enable Camera: Run sudo raspi-config → Interface Options → Camera → Enable
Connect Camera: Insert camera ribbon cable into the camera port (between Ethernet and HDMI)
Secure Connection: Make sure the ribbon cable is inserted correctly (metal contacts facing away from Ethernet port)
Reboot: Restart Raspberry Pi after enabling camera

Taking Pictures

The simplest thing you can do with a camera is take pictures!

Basic Photo Capture:

from picamera import PiCamera
import time

# Create camera object
camera = PiCamera()

# Optional: Set resolution
camera.resolution = (640, 480)

# Take a picture
camera.start_preview()  # Show preview
time.sleep(2)  # Give camera time to adjust
camera.capture('image.jpg')  # Save photo
camera.stop_preview()

print("Photo saved as image.jpg!")

Understanding the Code:

PiCamera(): Creates a camera object
start_preview(): Shows what camera sees (like a live view)
capture(): Takes a photo and saves it
stop_preview(): Closes the preview

Recording Video

You can also record videos with the camera:

Video Recording:

from picamera import PiCamera
import time

camera = PiCamera()
camera.resolution = (640, 480)

# Start recording
camera.start_recording('video.h264')
camera.wait_recording(10)  # Record for 10 seconds
camera.stop_recording()

print("Video saved as video.h264!")

Motion Detection

Motion detection compares consecutive frames to see if anything has moved. This is simpler than full object detection but very useful!

Simple Motion Detection:

from picamera import PiCamera
import numpy as np
import time

camera = PiCamera()
camera.resolution = (320, 240)  # Lower resolution for faster processing

def detect_motion():
    # Capture first frame
    camera.capture('frame1.jpg')
    time.sleep(0.1)
    
    # Capture second frame
    camera.capture('frame2.jpg')
    
    # Compare frames (simplified - in real project, you'd compare pixel differences)
    # If difference is large, motion detected!
    return True  # Simplified - always returns True for demo

try:
    while True:
        if detect_motion():
            print("Motion detected!")
            # Take a photo when motion detected
            camera.capture('motion_' + str(time.time()) + '.jpg')
        time.sleep(1)

except KeyboardInterrupt:
    camera.close()
    print("Motion detection stopped!")

Note: This is a simplified example. Real motion detection compares pixel differences between frames.

Project: Motion Detection Camera

Let's build a system that takes a photo when it detects motion!

Materials Needed:

Raspberry Pi
Raspberry Pi Camera Module
Optional: LED to indicate when motion is detected

Complete Motion Detection Code:

from picamera import PiCamera
import RPi.GPIO as GPIO
import time
from datetime import datetime

# Set up camera
camera = PiCamera()
camera.resolution = (640, 480)

# Set up LED indicator (optional)
GPIO.setmode(GPIO.BCM)
GPIO.setup(18, GPIO.OUT)

def take_motion_photo():
    """Take a photo when motion is detected"""
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    filename = f"motion_{timestamp}.jpg"
    
    # Flash LED to indicate photo taken
    GPIO.output(18, GPIO.HIGH)
    camera.capture(filename)
    GPIO.output(18, GPIO.LOW)
    
    print(f"Motion detected! Photo saved: {filename}")
    return filename

try:
    print("Motion detection camera started!")
    print("Press Ctrl+C to stop")
    
    last_capture_time = 0
    
    while True:
        # Simulate motion detection (in real project, compare frames)
        # For now, take a photo every 5 seconds as demo
        current_time = time.time()
        
        if current_time - last_capture_time >= 5:
            take_motion_photo()
            last_capture_time = current_time
        
        time.sleep(0.5)

except KeyboardInterrupt:
    camera.close()
    GPIO.cleanup()
    print("Motion detection stopped!")

Understanding the Code:

Camera continuously monitors for motion
When motion detected, takes a photo with timestamp
LED flashes to indicate photo was taken
Photos are saved with unique filenames

Introduction to Computer Vision

Computer vision is when computers "see" and understand images. It's what allows cameras to detect faces, objects, and people!

👁️

Image Capture

Camera takes pictures

🔍

Image Processing

Computer analyzes the image

🎯

Object Detection

Finds faces, people, objects

💡 Next Step: In the Advanced level, you'll learn to use OpenCV (a computer vision library) to detect faces and people - this is what you'll need for the human-tracking fan project!

Common Mistakes to Avoid

⚠️ Watch Out For:

Camera Not Enabled: Must enable camera in raspi-config first
Wrong Cable Orientation: Camera ribbon must be inserted correctly
Missing Library: Install picamera with pip3 install picamera
Too High Resolution: Higher resolution = slower processing - use lower for motion detection
Forgetting to Close: Always call camera.close() when done

Summary

You've learned:

✅ Raspberry Pi Camera can take photos and record videos
✅ Camera must be enabled in raspi-config before use
✅ Use PiCamera() to create camera object
✅ capture() takes photos, start_recording() records video
✅ Motion detection compares frames to detect movement
✅ Computer vision lets computers "see" and understand images
✅ Lower resolution = faster processing for motion detection

🎉 Congratulations! You've completed the Intermediate Level! You can now read sensors, display information, control motors, and use cameras. In the Advanced level, you'll combine all these skills to build the human-tracking fan!

🎮 Try It: Practice with Camera!

Practice writing code for the camera. Try these challenges:

📝 Challenge 1: Take a Photo

Write code that takes a single photo and saves it:

📝 Challenge 2: Timelapse

Write code that takes 10 photos, one every 2 seconds:

💡 Tip: Remember to enable the camera in raspi-config first! And always call camera.close() when you're done.

🎯 Activity: Motion Detection Camera

What You'll Build:

Create a motion detection system that takes photos when movement is detected!

Step-by-Step Instructions:

Enable Camera: Run sudo raspi-config and enable camera interface
Connect Camera: Insert camera ribbon cable into camera port
Install Library: Install picamera library if needed
Write Code: Create motion detection code
Test: Run code and wave in front of camera
Enhance: Add LED indicator when motion detected

Testing Checklist:

✅ Camera takes photos successfully
✅ Photos are saved with unique filenames
✅ System detects motion (or simulates detection)
✅ LED flashes when photo is taken

🏆 Bonus Challenge: Can you create a system that takes a photo every time a button is pressed? Combine camera with button input!

💪 Practice Challenges

Challenge 1: Photo with Timestamp

Take a photo and save it with the current date and time in the filename:

# Example filename: "photo_20241215_143022.jpg"
# (year, month, day, hour, minute, second)

Challenge 2: Continuous Photos

Take 5 photos, one every 3 seconds, and print a message each time:

Challenge 3: Button-Triggered Photo

Combine camera with button - take a photo when button is pressed:

# When button pressed → take photo
# Show LED indicator when photo taken

Challenge 4: Low-Resolution Motion Detection

Set camera to low resolution (160x120) for faster motion detection processing:

# Lower resolution = faster processing
# Better for real-time motion detection