视障辅助程序开发：如何快速分类RGB三元组至指定颜色组？

Hey there! No need to apologize at all—this is such a meaningful project, and your requirements are totally clear and practical. Let’s walk through a straightforward, high-fault-tolerance approach for the second step that checks all your boxes.

Unlike RGB (which is hardware-focused), HSV aligns much closer to how humans perceive color. The H (Hue) channel defines the actual color (red, blue, etc.), S (Saturation) measures vividness, and V (Value) controls brightness/darkness. This makes it perfect for:

• Fast, non-machine-learning execution
• Distinguishing light/dark shades of the same color
• Adapting to natural light brightness variations
• Easy tuning for high fault tolerance

Step 1: Convert RGB to HSV (Quick & Built-In) #

Most programming languages have built-in tools for this conversion—no external libraries needed. Here’s a Python example using the standard colorsys module:

import colorsys

def rgb_to_hsv(r, g, b):
    # Normalize RGB values from 0-255 to 0-1
    r_norm = r / 255.0
    g_norm = g / 255.0
    b_norm = b / 255.0
    # Convert to HSV (returns h: 0-1, s:0-1, v:0-1)
    h, s, v = colorsys.rgb_to_hsv(r_norm, g_norm, b_norm)
    # Convert hue to 0-360 for easier rule-writing
    return (h * 360, s, v)

Step 2: Define Color Group Rules (Covers All Required Categories) #

We’ll split rules into two groups: neutral colors (black, white, gray, beige) and colored shades (with light/dark distinctions). Rules are intentionally fuzzy to boost fault tolerance for natural light.

Neutral Color Rules #

Colored Shade Rules #

Each color includes thresholds to distinguish light/dark variants:

• Red: Hue 0-15 or 345-360
  • Dark (栗色): s > 0.3 AND v < 0.5
  • Bright Red: s > 0.3 AND v ≥ 0.5
  • Pink: s < 0.4 AND v ≥ 0.6 (low-saturation light red)
• Orange: Hue 15-45
  • Dark Orange: s > 0.25 AND v < 0.5
  • Bright Orange: s > 0.25 AND v ≥ 0.5
• Yellow: Hue 45-75
  • Dark Yellow: s > 0.2 AND v < 0.5
  • Bright Yellow: s > 0.2 AND v ≥ 0.5
• Green: Hue 75-160
  • Dark Green: s > 0.2 AND v < 0.5
  • Bright Green: s > 0.2 AND v ≥ 0.5
• Blue: Hue 160-260
  • Dark (藏青): s > 0.2 AND v < 0.5
  • Bright Blue: s > 0.2 AND v ≥ 0.5
• Purple: Hue 260-345
  • Dark Purple: s > 0.2 AND v < 0.5
  • Bright Purple: s > 0.2 AND v ≥ 0.5

Step 3: Full Implementation with Fault Tolerance #

Here’s a complete function that wraps everything up, with a fallback for edge cases (critical for accessibility):

import colorsys

def classify_color(r, g, b):
    h, s, v = rgb_to_hsv(r, g, b)

    # Handle neutral colors first
    if v < 0.2:
        return "黑色"
    elif v > 0.85 and s < 0.15:
        return "白色"
    elif 0.2 <= v <= 0.85 and s < 0.15:
        return "灰色"
    elif 0.7 <= v <= 0.95 and s < 0.3 and 20 <= h <= 60:
        return "米色"
    
    # Handle colored shades
    # Red/Pink family
    if (0 <= h <= 15) or (345 <= h <= 360):
        if s > 0.3:
            return "栗色（深红色）" if v < 0.5 else "红色"
        elif s < 0.4 and v >= 0.6:
            return "粉色"
    # Orange
    elif 15 < h <= 45:
        if s > 0.25:
            return "深橙色" if v < 0.5 else "橙色"
    # Yellow
    elif 45 < h <= 75:
        if s > 0.2:
            return "深黄色" if v < 0.5 else "黄色"
    # Green
    elif 75 < h <= 160:
        if s > 0.2:
            return "深绿色" if v < 0.5 else "绿色"
    # Blue
    elif 160 < h <= 260:
        if s > 0.2:
            return "藏青（深蓝色）" if v < 0.5 else "蓝色"
    # Purple
    elif 260 < h <= 345:
        if s > 0.2:
            return "深紫色" if v < 0.5 else "紫色"
    
    # Fallback for edge cases (critical for accessibility)
    return "灰色"

# Helper function from earlier
def rgb_to_hsv(r, g, b):
    r_norm = r / 255.0
    g_norm = g / 255.0
    b_norm = b / 255.0
    h, s, v = colorsys.rgb_to_hsv(r_norm, g_norm, b_norm)
    return (h * 360, s, v)

Step 4: Tuning Tips for Your Use Case #

Since you’re targeting natural light and high fault tolerance:

1. Test with real-world samples: Take photos of your target colors in sunny, cloudy, and indoor light, then adjust H/S/V thresholds if classifications are off.
2. Add majority voting: If your first step extracts multiple main colors (e.g., top 3), classify all and return the most frequent result—this reduces single-pixel errors.
3. Simplify voice output: For visually impaired users, use clear, concise labels (e.g., "深红色" instead of "栗色（深红色）" if preferred).

Hope this helps you get up and running quickly! This approach is beginner-friendly, fast, and designed to prioritize accessibility and accuracy.

内容的提问来源于stack exchange，提问作者Shahd Seddik