Object Detection with YOLO Algorithm Activity

What you'll learn

• How to install and configure the Ultralytics YOLO package
• Understanding YOLO model class names and detection capabilities
• Implementing real-time object detection on images
• Building object detection for video files
• Creating webcam-based detection applications
• Developing advanced object counting systems

Activity Overview

Total Time: 90-120 minutes

• Setup & Installation: 15 minutes
• Basic Detection: 60 minutes
• Advanced Challenge: 30-45 minutes

Activity Type: :computer: Coding & Implementation

Read, Think, Code

Question One: Setup and Installation ⏱️ 15 minutes

Objective: Install the YOLO package and verify installation

Steps to complete:

1. Open your terminal or command prompt
2. Navigate to your project directory
3. Run the appropriate installation command for your operating system:

• Windows:

py -m pip install -r requirements.txt

• MacOS:

python3 -m pip install -r requirements.txt

:bulb: Installation Success If successful, you should see a version number printed in your terminal (may vary depending on install date). If you encounter errors, make sure you have Python 3.7+ installed and pip is up to date.

Question 2: Understanding YOLO Model Classes ⏱️ 10 minutes

Objective: Explore what objects YOLO can detect

Steps to complete:

1. Create a new Python file called yolo_classes.py
2. Copy and run the following code:

from ultralytics import YOLO

# Load the trained model
model = YOLO('yolov8n.pt')

# Get class names
class_names = model.names

# Print class names
print(class_names)

# Save class names to file
with open('class_names.txt', 'w') as f:
    for index in class_names:
        f.write(f'{index}: {class_names[index]}\n')

3. Open the generated class_names.txt file to see all 80 object classes
4. Review the classes to understand what objects YOLO can detect

:information_source: About YOLO Classes YOLOv8 can identify 80 different object classes including people, animals, vehicles, and everyday objects. Learn more about the dataset at the YOLOv8 documentation.

Question 3: Detecting Objects in Images ⏱️ 20 minutes

Objective: Implement object detection on a static image

Steps to complete:

1. Place an image file named outing.jpg in your project folder
2. Create a new file called image_detection.py
3. Copy and run the following code:

from ultralytics import YOLO
import cv2
import math

# Load model
model = YOLO('yolov8n.pt')
class_name = model.names

# Load image
img = cv2.imread('outing.jpg')

# Detect objects
results = model(img)

# Draw detections
for r in results:
    for box in r.boxes:
        # Bounding box
        x1, y1, x2, y2 = box.xyxy[0]
        cv2.rectangle(img, (int(x1), int(y1)), (int(x2), int(y2)), (255, 0, 255), 3)
        
        # Confidence and class name
        conf = math.ceil((box.conf[0]*100))/100
        cls = int(box.cls[0])
        
        # Label
        label = f'{class_name[cls]}{conf}'
        cv2.putText(img, label, (int(x1), int(y1)-10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255), 2)

# Display image
cv2.imshow('Image', img)
cv2.waitKey(0)

4. Press any key to close the detection window

:bulb: Common Issues

• No detections? Try using an image with clearly visible objects like people, cars, or animals
• Low confidence scores? This is normal for small or partially visible objects
• Error loading image? Check that outing.jpg is in the same folder as your Python file

Question 4: Real-time Video Detection ⏱️ 20 minutes

Objective: Apply YOLO detection to video files

Steps to complete:

1. Place a video file named video.mp4 in your project folder
2. Create a new file called video_detection.py
3. Copy and run the following code:

from ultralytics import YOLO
import cv2
import math

# Load model
model = YOLO('yolov8n.pt')
class_name = model.names

# Load video
video = cv2.VideoCapture('video.mp4')

while True:
    ret, frame = video.read()
    if not ret:
        break
    
    # Detect objects
    results = model(frame)
    
    # Draw detections
    for r in results:
        for box in r.boxes:
            x1, y1, x2, y2 = box.xyxy[0]
            cv2.rectangle(frame, (int(x1), int(y1)), (int(x2), int(y2)), (255, 0, 255), 2)
            
            conf = math.ceil((box.conf[0]*100))/100
            cls = int(box.cls[0])
            label = f'{class_name[cls]}{conf}'
            cv2.putText(frame, label, (int(x1), int(y1)-10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
    
    cv2.imshow('Video', frame)
    if cv2.waitKey(1) == ord('q'):
        break

video.release()
cv2.destroyAllWindows()

4. Press 'q' to stop the video detection

:bulb: Performance Tips

• For faster processing, use smaller video files or lower resolution
• If the video runs slowly, try using yolov8s.pt instead of yolov8n.pt
• The first run may be slower as the model loads

Question 5: Live Webcam Detection ⏱️ 15 minutes

Objective: Create a real-time object detection system using your webcam

Steps to complete:

1. Ensure your webcam is connected and working
2. Create a new file called webcam_detection.py
3. Copy and run the following code:

from ultralytics import YOLO
import cv2
import math

# Load model
model = YOLO('yolov8n.pt')
class_name = model.names

# Initialize webcam
cap = cv2.VideoCapture(0)

while True:
    ret, frame = cap.read()
    
    # Detect objects
    results = model(frame)
    
    # Draw detections
    for r in results:
        for box in r.boxes:
            x1, y1, x2, y2 = box.xyxy[0]
            cv2.rectangle(frame, (int(x1), int(y1)), (int(x2), int(y2)), (255, 0, 255), 2)
            
            conf = math.ceil((box.conf[0]*100))/100
            cls = int(box.cls[0])
            label = f'{class_name[cls]}{conf}'
            cv2.putText(frame, label, (int(x1), int(y1)-10), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
    
    cv2.imshow('Webcam', frame)
    if cv2.waitKey(1) == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()

4. Press 'q' to stop the webcam detection

:warning: Privacy Notice This application uses your webcam. Make sure you're in a suitable environment before running this code. The detection runs locally on your computer - no data is sent anywhere.

Advanced Level ⏱️ 30-45 minutes

Advanced Challenge One: Building an Object Counter :dart:

Objective: Create a specialized counter for specific object types

Challenge Requirements:

• Count the number of people in an image
• Count the number of cats in an image
• Display the results clearly

Steps to complete:

1. Find or take a photo with both people and cats
2. Save it as advanced_image.jpg
3. Create advanced_counter.py with the following code:

# Advanced-1.py
from ultralytics import YOLO
import cv2

model = YOLO('yolov8n.pt')
img = cv2.imread('advanced_image.jpg')

# Class IDs for person and cat
person_id = list(model.names.keys())[list(model.names.values()).index('person')]
cat_id = list(model.names.keys())[list(model.names.values()).index('cat')]

results = model(img)

person_count = 0
cat_count = 0

for r in results:
    for box in r.boxes:
        cls = int(box.cls[0])
        if cls == person_id:
            person_count += 1
        elif cls == cat_id:
            cat_count += 1

print(f"People: {person_count}, Cats: {cat_count}")

:bulb: Extension Ideas

• Add more object types to count (dogs, cars, etc.)
• Display the counts on the image itself
• Create a GUI interface for selecting which objects to count
• Save detection results to a CSV file

Troubleshooting Guide

Common Issues and Solutions

1. Installation Errors

   • Update pip: pip install --upgrade pip
   • Try installing in a virtual environment
   • Check Python version (needs 3.7+)

2. Model Download Issues

   • The first run downloads the model (~6MB)
   • Ensure stable internet connection
   • If download fails, try again

3. Detection Not Working

   • Check image/video file paths
   • Verify webcam permissions
   • Try with different test images

4. Performance Problems

   • Use smaller model: yolov8n.pt (nano) is fastest
   • Reduce image/video resolution
   • Close other applications

Exercise Submission

:warning: Important: Submit Your Work! Submission Deadline: End of class

What to submit:

1. All your Python files (.py)
2. Screenshot of your detection results
3. Any modified code for the advanced challenge

Submit through this form:
:emoji: Exercise Submission Form

:information_source: Need Help? If you're stuck:

1. Review the error messages carefully
2. Check the troubleshooting guide above
3. Ask your instructor or classmates
4. Consult the YOLO documentation