- Function File:
*[*`level`,`sep`] =**graythresh***(*`img`) - Function File:
*[*`level`,`sep`] =**graythresh***(*`img`,`method`,`options`) - Function File:
*[*`level`,`sep`] =**graythresh***(*`hist`, …) Compute global image threshold.

Given an image

`img`finds the optimal threshold value`level`for conversion to a binary image with`im2bw`

. Color images are converted to grayscale before`level`is computed. An image histogram`hist`can also be used to allow for preprocessing of the histogram.The optional argument

`method`is the algorithm to be used (default’s to Otsu). Some methods may have other`options`and/or return an extra value`sep`(see each entry for details). The available`method`s are:- Otsu (default)
Implements Otsu’s method as described in Nobuyuki Otsu (1979). "A threshold selection method from gray-level histograms", IEEE Trans. Sys., Man., Cyber. 9 (1): 62-66. This algorithm chooses the threshold to minimize the intraclass variance of the black and white pixels.

The second output,

`sep`represents the “goodness” (or separability) of the threshold at`level`. It is a value within the range [0 1], the lower bound (zero) being attainable by, and only by, histograms having a single constant gray level, and the upper bound being attainable by, and only by, two-valued pictures.- concavity
Find a global threshold for a grayscale image by choosing the threshold to be in the shoulder of the histogram A. Rosenfeld, and P. De La Torre (1983). "Histogram concavity analysis as an aid in threshold selection", IEEE Transactions on Systems, Man, and Cybernetics, 13: 231-235.

- intermodes
This assumes a bimodal histogram and chooses the threshold to be the mean of the two peaks of the bimodal histogram J. M. S. Prewitt, and M. L. Mendelsohn (1966). "The analysis of cell images", Annals of the New York Academy of Sciences, 128: 1035-1053.

Images with histograms having extremely unequal peaks or a broad and flat valley are unsuitable for this method.

- intermeans
Iterative procedure based on the iterative intermeans algorithm of T. Ridler, and S. Calvard (1978). "Picture thresholding using an iterative selection method", IEEE Transactions on Systems, Man, and Cybernetics, 8: 630-632 and H. J. Trussell (1979). "Comments on ’Picture thresholding using an iterative selection method’", IEEE Transactions on Systems, Man, and Cybernetics, 9: 311.

Note that several implementations of this method exist. See the source code for details.

- MaxEntropy
Implements Kapur-Sahoo-Wong (Maximum Entropy) thresholding method based on the entropy of the image histogram J. N. Kapur, P. K. Sahoo, and A. C. K. Wong (1985). "A new method for gray-level picture thresholding using the entropy of the histogram", Graphical Models and Image Processing, 29(3): 273-285.

- MaxLikelihood
Find a global threshold for a grayscale image using the maximum likelihood via expectation maximization method A. P. Dempster, N. M. Laird, and D. B. Rubin (1977). "Maximum likelihood from incomplete data via the EM algorithm", Journal of the Royal Statistical Society, Series B, 39:1-38.

- mean
The mean intensity value. It is mostly used by other methods as a first guess threshold.

- MinError
An iterative implementation of Kittler and Illingworth’s Minimum Error thresholding J. Kittler, and J. Illingworth (1986). "Minimum error thresholding", Pattern recognition, 19: 41-47.

This implementation seems to converge more often than the original. Nevertheless, sometimes the algorithm does not converge to a solution. In that case a warning is displayed and defaults to the initial estimate of the mean method.

- minimum
This assumes a bimodal histogram and chooses the threshold to be in the valley of the bimodal histogram. This method is also known as the mode method J. M. S. Prewitt, and M. L. Mendelsohn (1966). "The analysis of cell images", Annals of the New York Academy of Sciences, 128: 1035-1053.

Images with histograms having extremely unequal peaks or a broad and flat valley are unsuitable for this method.

- moments
Find a global threshold for a grayscale image using moment preserving thresholding method W. Tsai (1985). "Moment-preserving thresholding: a new approach", Computer Vision, Graphics, and Image Processing, 29: 377-393

- percentile
Assumes a specific fraction of pixels (set at

`options`) to be background. If no value is given, assumes 0.5 (equal distribution of background and foreground) W Doyle (1962). "Operation useful for similarity-invariant pattern recognition", Journal of the Association for Computing Machinery 9: 259-267

**See also:**im2bw.

Package: image