How AI Sees the World 👁️

You See a Cat. The Computer Sees... Numbers?

When you look at a photo of a cat, you instantly think "cat!" 🐱

When a computer sees the same photo, it actually sees a giant grid of numbers:

Each pixel = 3 numbers (Red, Green, Blue values 0-255) [255, 200, 150] [230, 180, 120] [200, 160, 100] [240, 195, 140] [220, 175, 115] [195, 155, 95 ] [235, 190, 135] [215, 170, 110] [190, 150, 90 ] ...and so on for every single pixel! A 1000×1000 photo = 3,000,000 numbers!

Computer Vision is the AI technology that turns those millions of numbers back into understanding: "This is a cat. It's orange. It looks happy. It's outdoors."

🔍 How AI Looks at an Image (Step by Step)

Step 1: Look for EDGES Find where pixels change suddenly (That's where a cat's outline is!) Step 2: Look for SHAPES Combine edges into shapes (Pointed triangles = ears? Round shape = head?) Step 3: Look for PATTERNS Combine shapes into recognisable parts (Two triangular ears + round head + whiskers = cat face!) Step 4: Compare to TRAINING Match patterns to what it learned → "This matches CAT in training data!" 🐱 ✅

📸 Instagram Filters: Computer Vision in Action

When you use a Snapchat puppy filter or Instagram face filter:

1. Face detection: AI finds all faces in the camera feed (works even upside down!)
2. Landmark detection: AI identifies 68 specific points (corners of eyes, tip of nose, edges of mouth...)
3. Track in real time: Updates 30× per second as you move
4. Apply the filter: Dog ears follow your head, nose follows your nose!

All of this on your phone, in real time, FREE. Mind-blowing! 🤯

🌍 Computer Vision in the Real World

| Where | What it does | |-------|-------------| | 🚗 Self-driving cars | Detects pedestrians, signs, other cars | | 🏥 Hospitals | Analyses X-rays for cancer signs | | 🔒 Airports | Matches faces to passports | | 🛒 Amazon Go | Tracks everything you pick up (no checkout!) | | 🐘 Wildlife conservation | Counts endangered animals automatically |

💡 Did You Know?

Google Photos can find ALL your photos of a specific person — even across years and different haircuts — because of computer vision facial recognition. Try searching your name in Google Photos! 📸