Object Detection: YOLO Basics 📦

# Object Detection with YOLOv8 — Google Colab
# Install the Ultralytics library (includes YOLOv8)

!pip install ultralytics -q

from ultralytics import YOLO
import cv2
import numpy as np
from PIL import Image
import matplotlib.pyplot as plt
import requests
from io import BytesIO

# ── Step 1: Load a pre-trained YOLOv8 model ──
# yolov8n = nano (fastest, smallest) | yolov8s/m/l/x = larger, more accurate
model = YOLO('yolov8n.pt')   # downloads ~6MB model automatically
print("Model loaded! Classes:", len(model.names), "→", list(model.names.values())[:10], "...")

# ── Step 2: Run detection on an image URL ──
# Using a public traffic image (replace with your own photo path)
img_url = "https://upload.wikimedia.org/wikipedia/commons/thumb/4/47/PNG_transparency_demonstration_1.png/280px-PNG_transparency_demonstration_1.png"

# For a local file: results = model.predict("path/to/image.jpg", conf=0.5)
# For Google Drive: results = model.predict("/content/drive/MyDrive/my_image.jpg")
results = model.predict(source=img_url, conf=0.5, iou=0.5)

# ── Step 3: Inspect raw results ──
result = results[0]
print(f"\nDetections found: {len(result.boxes)}")
for box in result.boxes:
    cls_id = int(box.cls)
    label  = model.names[cls_id]
    conf   = float(box.conf)
    x1, y1, x2, y2 = box.xyxy[0].tolist()  # absolute pixel coords
    print(f"  {label:15s}  conf={conf:.2f}  box=[{x1:.0f},{y1:.0f},{x2:.0f},{y2:.0f}]")

# ── Step 4: Visualise with bounding boxes ──
annotated = result.plot()   # returns numpy array with boxes drawn
plt.figure(figsize=(10, 6))
plt.imshow(cv2.cvtColor(annotated, cv2.COLOR_BGR2RGB))
plt.axis('off')
plt.title('YOLOv8 Detections')
plt.show()

# ── Step 5: Count objects by class ──
from collections import Counter
detected_classes = [model.names[int(b.cls)] for b in result.boxes]
counts = Counter(detected_classes)
print("\nObject counts:", dict(counts))

# ── Step 6: Apply to your own image ──
def detect_and_show(image_path, conf_threshold=0.5):
    """Detect objects in any image and display results."""
    results = model.predict(source=image_path, conf=conf_threshold)
    r = results[0]
    annotated = r.plot()
    plt.figure(figsize=(12, 7))
    plt.imshow(cv2.cvtColor(annotated, cv2.COLOR_BGR2RGB))
    plt.axis('off')
    plt.title(f"Detections (conf ≥ {conf_threshold}): {len(r.boxes)} objects")
    plt.show()
    for box in r.boxes:
        print(f"  {model.names[int(box.cls)]:20s}  conf={float(box.conf):.3f}")

# Test with your own uploaded image:
# from google.colab import files
# uploaded = files.upload()
# detect_and_show(list(uploaded.keys())[0])

# ── Step 7: Process a video or webcam (advanced) ──
# results = model.predict(source="traffic.mp4", show=True, stream=True)
# for r in results:
#     ... process each frame ...