What it does
This KPI predicts the perceived brightness of an image. Rather than measuring objective brightness, it calculates brightness as processed by human vision.
Why it matters
Brightness plays a critical role in creating visual interest and dynamic contrast in marketing assets. It ensures that assets grab attention and stand out in different environments, which is especially important for out-of-home and mobile asset consumption where lighting conditions can vary significantly. High perceived brightness makes assets more visible and impactful whether they are viewed in bright sunlight, a well-lit room, or a darker environment.
How it works
Using the Hue, Saturation, Perceived Brightness (HSP) color model, which accounts for the human perception of color intensity, the average brightness is calculated from the RGB values of an image. The perceived brightness for each color is weighted according to how the human eye perceives brightness, giving more weight to colors that appear brighter to the human eye.
Examples: Black has a perceived brightness of 0 while white as a score of 255 (the maximum). For the blue tones we see that the light blue (206) has a much higher brightness than the dark blue (47).
How to achieve great results
The illustration shows sample images with brigthness predictions (scaled to range [0...100]). Top left image has lowest, bottom right image has highest perceived image brightness.
High brightness drivers:
- Bright Colors and Light Backgrounds: Use of white, pastels, or bright colors increases perceived brightness.
- High Contrast: Strong contrast between elements creates a brighter visual.
- Direct Lighting: Emphasis on well-lit subjects enhances overall brightness.
Low brightness drivers:
- Dark Colors: Predominantly dark color schemes reduce perceived brightness.
- Muted Tones and Shadows: Muted colors and heavy shadows contribute to lower brightness.
- Low Contrast: Lack of distinct contrast leads to a dimmer perception.
AI models used
A proprietary algorithm calculates brightness levels based on how they are perceived by the human eye. The algorithm reflects the non-linear nature of human brightness perception, ensuring a realistic and effective assessment of image brightness in assets.