Local adaptive thresholding with GUI to help find best parameters.

Local adaptive thresholding based on local mean and standard deviation.

Local adaptive thresholding : using local mean and standard deviation for foreground segmentation.
Two usage:
Using GUI to find best parameter and perform thresholding.

[ bw, params ] = localthreshGUI(img) takes img and displays it and segmented image.

Upon existing GUI program, binary image (bw) and parameters for localthresholding (params) are returned.

Using the function without GUI.

bw = localthresh(img, ssize, msize, sthresh, mthresh) performs local threshing holding on image.

ssize : filter size for standard deviation filter. msize : filter size for mean filter sthresh : threshold for standard deviation ( in fold number relative to local standard deviation ) mthresh : threshold for mean ( in fold number relative to local mean )

Type in the following to try it :

%load built-in rice image
img = imread('rice.png')
localthreshGUI(img)