← Image Processing for EveryoneToolsInVisionLearning Center
Histograms and Color · Lesson 28

White Balance Under Changing Light

The same leaf looks warmer at sunset than at noon.

Watch the lesson

White Balance Under Changing Light explained visually

1:28

The same leaf looks warmer at sunset than at noon. English captions are optional and start off. The complete transcript follows the written lesson.

Lesson overview

What you will learn

The same leaf looks warmer at sunset than at noon.

  1. A practical question
  2. The useful idea
  3. Work through the example
  4. Limits and next step
01

A practical question

The same green leaf looks warmer in a sunset photo than it did at noon. Some of that color difference comes from the light, not the leaf. White balance tries to compensate for the illumination so that a neutral object appears neutral in the image. We need an actual reference or a justified estimate, not simply a slider position that makes one leaf look pleasing.

02

The useful idea

Photograph a gray card under each lighting setup. If one light pushes the card toward orange, the channel response gives us a clue about the correction needed for that capture. Apply a correction derived for that light, and compare the card and leaf afterward. A setting from noon should not automatically be reused at sunset.

03

Work through the example

This also has limits. A card under one lamp may not describe a leaf partly in shadow and partly in sun. Reflections can change individual regions too. Keep the originals and the reference images, and inspect several leaves rather than claiming that one neutral patch makes every surface perfectly comparable.

04

Limits and next step

For our analysis, the goal is consistent evidence about the leaves, not the most vivid photograph. Next, we'll return to spatial filtering and build a neighborhood operation from its weights.

Worked example

Compare a gray card under both lights and adjust each image from its own reference.

Knowledge check

Quick check: test your understanding

Try it yourself: In the situation described above, explain the method and one reason its output might be misleading.

Check the reasoning

White balance compensates for the illuminant using channel gains or a reference neutral. A neutral card and white balance do not remove every shadow or reflection.

What is the main idea behind white balance under changing light?

White balance compensates for the illuminant using channel gains or a reference neutral.

How is it used in the example?

Compare a gray card under both lights and adjust each image from its own reference.

What limitation should be checked?

A neutral card and white balance do not remove every shadow or reflection.

Further reading

Digital Image Processing, 4th edition — Rafael C. Gonzalez and Richard E. Woods.

Original ToolsInVision teaching examples. Editorial approach · Suggest a correction

Accessible referenceLesson 28 transcript200 words · Video transcript · Matches the published narration and captionsOpen transcript

The same green leaf looks warmer in a sunset photo than it did at noon. Some of that color difference comes from the light, not the leaf. White balance tries to compensate for the illumination so that a neutral object appears neutral in the image. We need an actual reference or a justified estimate, not simply a slider position that makes one leaf look pleasing.

Photograph a gray card under each lighting setup. If one light pushes the card toward orange, the channel response gives us a clue about the correction needed for that capture. Apply a correction derived for that light, and compare the card and leaf afterward. A setting from noon should not automatically be reused at sunset.

This also has limits. A card under one lamp may not describe a leaf partly in shadow and partly in sun. Reflections can change individual regions too. Keep the originals and the reference images, and inspect several leaves rather than claiming that one neutral patch makes every surface perfectly comparable.

For our analysis, the goal is consistent evidence about the leaves, not the most vivid photograph. Next, we'll return to spatial filtering and build a neighborhood operation from its weights.