Image Processing 101 Chapter 1.2: Color Models

Jul 16, 2025 · Alok
  Image-processing

What is a Color Model?

A color model is a mathematical way to describe and represent colors using numbers – typically in tuple like RGB (three values: Red, Green, Blue) or quadruple like CMYK (four values: Cyan, Magenta, Yellow, Black). These values define positions in a coordinate system that helps us encode color information digitally.

For a given color model, to interpret a tuple or quadruple as a color, we can define a set of rules and definitions used to accurately calibrate and generate colors, i.e. a mapping function. A color space identifies a specific combination of color models and mapping functions. Identifying the color space automatically identifies the associated color model. For example, Adobe RGB and sRGB are two different color spaces, both based on the RGB color model.

Common Color Models

RGB (Red, Green, Blue)

RGB color model stores individual values for red, green, and blue. It is an additive model. With a color space based on the RGB color model, the three primaries are added together to create colors from completely white to completely black.

The RGB color space is associated with the device. Thus, different scanners get different color image data when scanning the same image; different monitors have different color display results when rendering the same image.

RGB Color Model

Color Correction and Enhancement Using RGB Model - Basic Color Balance and Contrast Adjustment in RGB

Here is a photo taken in the shade with a strong blue tint. To balance the color, we increased red and green channels slightly to balance out the blue and Contrast adjusted by using channel curves.

rgb-color-balance

HSV /HSB (Hue, Saturation, Value/Brightness)

HSV (hue, saturation, value), also known as HSB (hue, saturation, brightness), is often used by artists because it is often more natural to think about a color in terms of hue and saturation than in terms of additive or subtractive color components.

The system is closer to people’s experience and perception of color than RGB. For example, in painting terms, hue, saturation, and values are expressed in terms of color, shading, and toning.

The HSV space is often visualized as a hexagonal cone or cylinder where Hue changes around the circle, Saturation increases from center outward and Value increases from bottom to top. It allows for intuitive color selection which is useful in segmentation, tracking, and filtering where color needs to be isolated from brightness.

HSV Color Model

Color Correction and Enhancement Using HSV Model - Changing Hue and Saturation in An Image

In this example, we are making HSV-based enhancements in a Sunset color. The first photo shows a landscape with dull green grass and faded sky. To enhance the image, we shifted Hue slightly to make the sky more orange and boosted Saturation for grass to appear more vivid and increased Value for overall brightness.

HSV-adjustment

YUV/Y’UV – Luma-Chroma Separation

The Y′UV model defines a color space in terms of one luma component (Y’) and two chrominance (UV) components. The Y’ channel carries grayscale brightness information and is computed using a weighted sum of RGB values (Y’=0.299R +0.587G+0.114B). The U and V components encode color differences – U represents the blue projection, and V the red projection relative to the luma. This separation is particularly effective for video applications because the human eye is more sensitive to brightness than color. By storing color data at a lower resolution, it becomes possible to compress images and videos significantly without visibly degrading quality.

YUV Color Model

Color Correction and Enhancement Using YUV Model - Contrast Adjustment Without Affecting Color

This use-case shows how luminance correction improves visibility without making colors look unnatural. The original image was shot indoors with low-light, and it looks dull and flat. To make the enhancement, Y(Luma) contrast was stretched, Cb/Cr (color components) we kept unchanged to preserve the colors.

YUV-adjustment

Application in Object Recognition and Segmentation

Color models play a pivotal role in object recognition and image segmentation. Models like HSV and YUV are preferred over RGB for such tasks due to their better separation of luminance and chromatic data.

In object detection, HSV is widely used because it enables robust color thresholding under varying lighting conditions – by isolating the hue, objects of the same color can be detected even if their brightness differs. For example, red traffic signs can be isolated by selecting a hue range near 0o regardless of the lighting.

YUV on the other hand, allows for efficient edge detection and object tracking by using the Y (luma) channel alone. This simplifies computations without compromising object outlines.

The Lab color model is also used in segmentation, especially when perceptual accuracy is important. Its separation of lightness from color components and perceptual uniformity make it effective for clustering-based segmentation methods. The ‘a’ and ‘b’ channels encode color in a way that aligns with human vision, helping algorithms detect subtle color differences more reliably. Lab models are often used in preprocessing for machine learning pipelines where consistent color-based features improve model performance.

Conclusion

Color models are fundamental to how we represent, manipulate, and analyze images in both simple and advanced image processing workflows. Each model serves a unique purpose. Understanding when and why to use a specific model allows developers and researchers to build more accurate, efficient, and visually coherent image processing pipelines. Whether you’re enhancing image quality, performing color correction, or training neural network to detect objects, choosing the right color model can make a significant difference in outcome and performance.

To learn more about imaging fundamentals, read the first article in the Image Processing 101 Series: What is an Image. Subscribe to our newsletter and learn more about Image Processing.

Reference: Wikipedia

This article is Part 2 in a 7-Part Series.