Otsu's method

Last revised by Joachim Feger on 7 May 2024

Otsu’s method, also known as Otsu's threshold algorithm, is an automatic thresholding algorithm used in image processing.

The method separates pixels or voxels from an image into groups based on where their values lie in relation to the image histogram. The method finds threshold values that create a separation to minimize intraclass pixel/voxel value variance or conversely, maximize the interclass variance of values.

Many image processing libraries have built-in implementations (pre-made functions) of Otsu’s method that will return the threshold value computed with this method and there are many variations and extensions of Otsu’s method that have been applied to radiological images 2,3.

ADVERTISEMENT: Supporters see fewer/no ads

Updating… Please wait.

 Unable to process the form. Check for errors and try again.

 Thank you for updating your details.