| Conversion | Output Channels | Key Properties |
|---|---|---|
| BGR ↔ HLS | Hue (0–360°), Lightness (0–1), Saturation (0–1) | Lightness = average of max and min RGB channels. Saturation varies with lightness. |
| BGR ↔ HSV | Hue (0–360°), Saturation (0–1), Value (0–1) | Value = max RGB channel. More intuitive than HLS for image processing. Hue undefined when S=0 (grayscale). |
| BGR ↔ Lab | L* (0–100), a* (green↔red), b* (blue↔yellow) | Perceptually uniform. L* channel is close to perceived lightness. a* and b* encode color information orthogonally. |
| RGB → Grayscale | Single intensity channel | Weighted average: I = 0.299R + 0.587G + 0.114B (NTSC weights). Non-invertible. Green-channel weighted because human vision is most sensitive to green. |
| Convert BGR ↔ HLS | Bidirectional | Convert to HLS for analysis, back to BGR for visualization. |
| Convert BGR ↔ HSV | Bidirectional | Most common color space conversion for image processing tasks. |
| Convert BGR ↔ Lab | Bidirectional | Best for color comparison and perceptual analysis. |
BOM Group
Color Space Conversions
Transforming between color representations
View
Definition
Operations for converting images between color representations — BGR↔HLS, BGR↔HSV, BGR↔Lab, and RGB to Grayscale — separating chromatic information from intensity for analysis tasks where hue and brightness need independent processing.
Four Color Spaces
HLS, HSV, Lab, and Grayscale
Separate Color from Brightness
Hue identifies stain type regardless of staining strength
Perceptual Uniformity (Lab)
Equal numerical distance = equal perceived difference
Non-Invertible Grayscale
RGB → Gray loses color information permanently
Operations Reference
Simplified
These conversions transform images from raw camera RGB into color spaces that separate color from brightness. HSV is most commonly used — Hue tells you the color, Saturation tells you how vivid it is, Value tells you how bright. Lab is best for comparing colors because equal distances mean equal visual differences. Grayscale reduces to a single brightness channel.
When to Use Which
HSV for Stain Type Identification
In brightfield imaging, different chromogens (DAB brown, hematoxylin blue, Fast Red) have distinct hues. Converting to HSV and thresholding on the hue channel can separate stain types by color alone, regardless of staining intensity. This is a simpler alternative to full color deconvolution for some applications.
Lab for Color Distance
When comparing stain colors — e.g., determining whether two reference shades represent the same chromogen — Lab provides the most meaningful distance metric. Two colors with ΔE (Euclidean distance in Lab) < 2 are visually indistinguishable.
Grayscale for Single-Channel Processing
Many BOM operations (filtering, thresholding, morphology) require grayscale input. RGB to Grayscale converts color images to a single intensity channel using perceptually weighted averaging. Note: for quantitative IHC analysis, use Color Separation or OD conversion instead — they provide stain-specific channels rather than a generic intensity average.
Histogram Equalization in HSV
Solomon & Breckon warn that equalizing R, G, B channels independently "will alter the chromatic (colour hue) content." The correct approach: convert to HSV, equalize only the V channel, then convert back — this enhances contrast without distorting colors.
Simplified
Use HSV when you need to separate color from brightness — e.g., finding all purple-stained regions regardless of staining strength. Use Lab when comparing colors quantitatively. Use Grayscale when you need a single intensity channel for operations that don't understand color. For quantitative IHC analysis, prefer Color Separation or OD conversion over simple grayscale.