Federal Agencies
Digitization Guidelines Initiative

Home > Glossary > T > Thresholding

Term: Thresholding

 “Search Glossary” button searches only the glossary. Temporary note: search not enabled for two- and three-character terms; browse by alphabet.
 “Search“ button at the top right of the page searches the Web site, not the glossary.

Suggest a term

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

Term: Thresholding

An image processing method that creates a bitonal (aka binary) image based on setting a threshold value on the pixel intensity of the original image. While most commonly applied to grayscale images, it can also be applied to color images. The threshold of image intensity (relative image lightness) is set manually at a specific value or automatically set by an application. Pixels below that set threshold value are converted to black (bit value of zero), and pixels above the threshold value are converted to white (a bit value of one). The thresholding process is sometimes described as separating an image into foreground values (black) and background values (white).

Simple thresholding operations establish a single global threshold value for all pixels in an image irrespective of any local variations in contrast. More sophisticated thresholding processes (adaptive thresholding) sample small and greater number of regions of the image and establish the threshold value accordingly.

The quality of thresholding is of particular importance when performing OCR on images. Foxed, mottled, stained or irregularly faded materials can present a challenge in properly separating foreground (text) from background.
See also:
Bitonal digital image; Thresholding, adaptive