Lesson 26: Color Recognition & Computer Vision

1. Understanding Color and Computer Vision

Color recognition allows robots to identify, track, and respond to colored objects. This enables applications like traffic light recognition, object sorting, and color-based navigation.

2. ESP32-S3 Camera Programming

Basic Color Detection:

#include "esp_camera.h"

// Color range structure
struct ColorRange {
  uint8_t r_min, r_max;
  uint8_t g_min, g_max;
  uint8_t b_min, b_max;
};

// Red color range
ColorRange redRange = {150, 255, 0, 100, 0, 100};

bool detectColor(uint8_t r, uint8_t g, uint8_t b, ColorRange range) {
  return (r >= range.r_min && r <= range.r_max &&
          g >= range.g_min && g <= range.g_max &&
          b >= range.b_min && b <= range.b_max);
}

Color Tracking Algorithm:

void followColorObject() {
  camera_fb_t * fb = esp_camera_fb_get();
  if (!fb) return;
  
  uint16_t* pixels = (uint16_t*)fb->buf;
  int sumX = 0, sumY = 0, count = 0;
  
  // Find color centroid
  for (int y = 0; y < fb->height; y++) {
    for (int x = 0; x < fb->width; x++) {
      int index = y * fb->width + x;
      uint16_t pixel = pixels[index];
      
      uint8_t r = (pixel >> 11) << 3;
      uint8_t g = ((pixel >> 5) & 0x3F) << 2;
      uint8_t b = (pixel & 0x1F) << 3;
      
      if (detectColor(r, g, b, redRange)) {
        sumX += x;
        sumY += y;
        count++;
      }
    }
  }
  
  if (count > 50) {
    int centerX = sumX / count;
    int error = centerX - 160; // Image center
    
    // Control robot based on object position
    if (abs(error) < 20) {
      moveForward(100);
    } else if (error > 0) {
      turnRight(50);
    } else {
      turnLeft(50);
    }
  }
  
  esp_camera_fb_return(fb);
}

3. Practical Applications

Build a robot that detects, tracks, and follows colored objects using the ESP32-S3 camera.

Quick Assessment

• ✅ Explain RGB vs HSV color spaces
• ✅ How does ESP32-S3 detect colors?
• ✅ What affects color recognition accuracy?
• ✅ Design a color-based sorting system

Homework Assignment

• 📝 Research real-world color recognition applications
• 🎨 Test different lighting conditions on color detection
• 🤖 Program a simple traffic light response system
• 📊 Document which colors work best/worst and why