RATS_

RATS_ (Robust Automatic Threshold Selection)

Author: Ben Tupper (btupper at bigelow.org) and Mike Sieracki (msieracki at bigelow.org)

History: 2009-01-08: First version
2009-12-11: BTT, minor fixes

Source: RATS_.jar

Installation: Download RATS_.jar to the plugins folder or a subfolder and restart ImageJ or use the menu selection Help > Update Menus.

Description: RATS_ (Robust Automatic Threshold Selection) is based upon work of M.H.F. Wilkinson and others, RATS establishes regionalized thresholds for a greyscale image where the regions are established using recursive quadtree architecture. Within each of the lowest quadtree regions the threshold is calculated as the sum of the orginal pixels weighted by the gradient pixels.
for pixels within each region. The threshold calculated in each region is required to meet minimum criteria - these criteria are determined by the user as a noise estimate (sigma) and a scaling factor (lambda). The user may also select the minimum region size. The thresholds for all region are then interpolated (bilinear) across the entire image. In general, the best values for each of three parameters are determined by trial and error for a given suite of images. See references in source code and below.
Source
REFERENCE: M.H.F. Wilkinson (1998) Optimizing edge detectors for robust automatic threshold selection. Graph. Models Image Proc. 60: M.H.F. Wilkinson (1996) Rapid automatic segmentation of fluorescent and phase-contrast images of bacteria. In: Fluorescence Microscopy and Fluorescent Probes,(J. Slavik, ed), pp 261-266, Plenum Press, New York.
REFERENCE: M.H.F. Wilkinson "Segmentation Techniques in Image Analysis of Microbes" which is chapter 3 in Digital Image Analysis of Microbes: Imaging, Morphometry, Fluorometry and Motility Techniques and Applications. Edited by M.H.F. Wilkinson and F. Schut, Wiley Modern Microbiology Methods Series (1998).

The left image shows the dialog in which the user is prompted for...
1. NOISE THRESHOLD: An estimate of the noise. Estimate the noise by selecting a "background" portion of the image and using ImageJ to determine the standard deviation of gray values. Oddly, lower values yield smaller particles in general. (see refs, defaults to 25).
2. LAMBDA FACTOR: A scaling factor. Higher values yield larger particles. (see refs, defaults to 3)
3. MIN LEAF SIZE (pixels): The smallest allowed leaflet (defaults to 5 levels of quadtrees so the default value is computed on the fly based upon the input image width and height.
4. VERBOSE If set then output informational messages in the log window (default is false).

The center and right images show an example image its edge enhancement with the nested quadtrees for just one quadrant superimposed.

|Plugins | Home |

Author:	Ben Tupper (btupper at bigelow.org) and Mike Sieracki (msieracki at bigelow.org)
History:	2009-01-08: First version 2009-12-11: BTT, minor fixes
Source:	RATS_.jar
Installation:	Download RATS_.jar to the plugins folder or a subfolder and restart ImageJ or use the menu selection Help > Update Menus.
Description:	RATS_ (Robust Automatic Threshold Selection) is based upon work of M.H.F. Wilkinson and others, RATS establishes regionalized thresholds for a greyscale image where the regions are established using recursive quadtree architecture. Within each of the lowest quadtree regions the threshold is calculated as the sum of the orginal pixels weighted by the gradient pixels. for pixels within each region. The threshold calculated in each region is required to meet minimum criteria - these criteria are determined by the user as a noise estimate (sigma) and a scaling factor (lambda). The user may also select the minimum region size. The thresholds for all region are then interpolated (bilinear) across the entire image. In general, the best values for each of three parameters are determined by trial and error for a given suite of images. See references in source code and below. Source REFERENCE: M.H.F. Wilkinson (1998) Optimizing edge detectors for robust automatic threshold selection. Graph. Models Image Proc. 60: M.H.F. Wilkinson (1996) Rapid automatic segmentation of fluorescent and phase-contrast images of bacteria. In: Fluorescence Microscopy and Fluorescent Probes,(J. Slavik, ed), pp 261-266, Plenum Press, New York. REFERENCE: M.H.F. Wilkinson "Segmentation Techniques in Image Analysis of Microbes" which is chapter 3 in Digital Image Analysis of Microbes: Imaging, Morphometry, Fluorometry and Motility Techniques and Applications. Edited by M.H.F. Wilkinson and F. Schut, Wiley Modern Microbiology Methods Series (1998). The left image shows the dialog in which the user is prompted for... 1. NOISE THRESHOLD: An estimate of the noise. Estimate the noise by selecting a "background" portion of the image and using ImageJ to determine the standard deviation of gray values. Oddly, lower values yield smaller particles in general. (see refs, defaults to 25). 2. LAMBDA FACTOR: A scaling factor. Higher values yield larger particles. (see refs, defaults to 3) 3. MIN LEAF SIZE (pixels): The smallest allowed leaflet (defaults to 5 levels of quadtrees so the default value is computed on the fly based upon the input image width and height. 4. VERBOSE If set then output informational messages in the log window (default is false). The center and right images show an example image its edge enhancement with the nested quadtrees for just one quadrant superimposed.