🎨 Colour Spaces: RGB, HSV, YCrCb

Colour spaces define how colours are encoded and interpreted in digital images. Choosing the right space can simplify tasks like segmentation, tracking and compression.

---

🔍 What is a Colour Space?

A colour space is a particular organisation of colours, allowing us to represent them as numerical values. Different spaces emphasise different aspects of colour, making them useful for various computer-vision tasks.

RGB = mixing paint 🎨
HSV = describing colour in human terms (hue, saturation, brightness) 🌈
YCrCb = separating brightness (luma) from colour (chroma) 🎥

---

🔵 RGB (Red, Green, Blue)

• The most common colour space.
• Additive colour model: combine red, green and blue to create other hues.
• Used by screens, cameras and OpenCV by default.

Example pixel:

[R, G, B] = [120, 60, 200]

🌈 HSV (Hue, Saturation, Value)

Closer to how humans perceive colour.
Very useful for colour-based filtering (e.g. detecting red objects).

Components

Component	Description
Hue (H)	Colour type (0–179 in OpenCV, e.g. red, blue)
Saturation (S)	Intensity/purity of colour (0–255)
Value (V)	Brightness (0–255)

Example:

Pure red:  [H=0, S=255, V=255]  
Light red: [H=0, S=100, V=255]  

🎥 YCrCb (Luminance + Chrominance)

Separates Y (luminance/brightness) from Cr and Cb (red and blue chroma).

Widely used in video and image compression (JPEG, MPEG).

Component	Meaning
Y	Luminance (brightness)
Cr	Red-difference chroma
Cb	Blue-difference chroma

Advantage: Brightness can be processed separately, improving compression and lighting control.

🧪 Python Code: RGB → HSV and YCrCb Conversions

import cv2

# Read image in colour
img = cv2.imread("bird.jpg")

# Convert to HSV
img_hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)

# Convert to YCrCb
img_ycrcb = cv2.cvtColor(img, cv2.COLOR_BGR2YCrCb)

# Check shape and types
print("Original (BGR):", img.shape)
print("HSV:", img_hsv.shape)
print("YCrCb:", img_ycrcb.shape)

Output:

Original (BGR): (144, 349, 3)
HSV: (144, 349, 3)
YCrCb: (144, 349, 3)

All conversions keep the same dimensions and three channels.

🔍 Visualization Tip

To see how the colour-space conversions affect the image, plot all three side-by-side using Matplotlib:

import cv2
import matplotlib.pyplot as plt

# Read the image
img = cv2.imread("bird.jpg")
img_hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
img_ycrcb = cv2.cvtColor(img, cv2.COLOR_BGR2YCrCb)

# Convert BGR → RGB for correct display in Matplotlib
img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

# Plot
plt.figure(figsize=(15, 5))

plt.subplot(1, 3, 1)
plt.imshow(img_rgb)
plt.title("Original (RGB)")
plt.axis("off")

plt.subplot(1, 3, 2)
plt.imshow(img_hsv)
plt.title("HSV")
plt.axis("off")

plt.subplot(1, 3, 3)
plt.imshow(img_ycrcb)
plt.title("YCrCb")
plt.axis("off")

plt.tight_layout()
plt.show()

✅ Summary Table

Colour Space	Description	Use Case
RGB	Red, Green, Blue	Default format
HSV	Hue, Saturation, Value	Colour filtering, segmentation
YCrCb	Luminance + Chrominance	Compression, illumination