Image Histogram

An image histogram is a type of statistical graph distribution of color intensities in an image. It uses the abscissa of the colors and the number of pixels for each color for the ordinate, which makes it possible to quickly evaluate the distribution of the colors within the image image.

In practice, the colors are usually limited to 255 levels of gray, the image is previously converted into grayscale (luminance).

Iterate over each pixel in the image and calculate the gray level of the pixel. Increment a counter corresponding to each gray level.

The histogram is the summary table of these values, which can then be displayed in graphic form.

On dCode, this work is done automatically, upload an image file (JPG, PNG, etc.) and click on Analyze.

On the horizontal axis are the 256 gray/luminance values.

On the vertical axis the number of pixels in the image having this luminance value.

A histogram can be used to determine whether an image is underexposed or overexposed, in order to adjust the brightness and contrast of the image.

It is potentially possible to hide information in the histogram, such as a drawing/silhouette.

Histogram is based on the frequency of luminance in the image from 0 (black) to 255 (white). The luminance is computed for each pixel with the formula 0.2126*R+0.7152*G+0.0722*B (ITU BT.709).

NB: to speed up the calculations, sometimes the formula (R + G + B) / 3 is used (not recommended)

The colors are organized according to their tone, blacks are on the left and on the right are whites, in the middle are gray tones.

Gray levels or color values are usually normalized to a scale of 0 to 255.

Use the imread function to load an image, then imhist to display its histogram.

To assess contrast from a histogram, observe the distribution of values along the horizontal axis.

A histogram with a distribution spread over the entire range of grayscale or color indicates good contrast, while a histogram concentrated in a narrow range may indicate poor contrast.

An overexposed image will tend to have a histogram pushed to the right, with an accumulation of pixels towards higher grayscales or color values.

An underexposed image will show a shifted histogram to the left, with pixels mostly concentrated towards grayscale or low color values.

Lumens are a measure of the total amount of visible light emitted by a light source. A histogram does not directly calculate lumens, but only the distribution of light intensities in an image. If this image is (back-)lit, then it is the power of this lighting which will provide lumens.