Macro Single Folder - Instructions

BASIC OPERATIONS
To operate this macro you need:
- ImageJ v. 1.38a or higher.
- the file of the macro as a .txt text file that you can obtain clicking on the <Download macro> link of htis site.
- a set of images grouped in a folder.

The images must be black and white (thresholded) images with white background. All of the images must be perfect squares with exactly the same dimensions, in order to obtain meaningful results.

Upon starting the macro you will presented with a file opening window, and the macro will ask you:
- where are the images to be analyzed.
- where you wish to save the result file.
After that, the macro will run autonomously and can be leaved unattended. During the elaboration a progress bar will show the percent of the images that have been processed. please note that the elaboration does not affect the input images that can be reutilized at will.
At the end, a message will inform that the job has been completed. The results are available, in the position choseen, as a file with the name of the analyzed folder.

STRATEGY FOR MULTIPLE COMPARISONS
If you need to compare different sets of images grouped in separate folders you should normalize ALL the images against the one showing a maximal value. This task can be accomplished in 2 steps:
- first: identifying the image with the largest number of pixels using the utility < GetInitialAreas > (an ImageJ macro itself) that you find in the Toolbox section of this site.
- second: copying the above image in every folder you wish to compare. It is important that you take note of its name in order to remove its data from the result file.
At this point you might be willing to use the Multi Folders Macro, otherwise you will perform n analysis, a folder at a time.

MACRO LOGIC
The quantification of the space-filling property of a layout of elements is carried out in 2 steps.
- First step: the macro calculates the initial percent of black pixels for every image of the set and identifies the image showing the largest value. This value is taken as the normalization value to derive normalized H-indexes.
- Second step: for every image of the set, taken in alphabetical order, the macro performs pixel dilations up to a maximum of 500 cycles or until the dilated area exceedes 99% of the image. At this point the macro calculates how many cycles were necessary to reach the normalization value (N) and the cycles necessary to cover 90%, 95% and 99% of the image. These values are recorded as the raw H indexes. Normalized H indexes are obtained subtracting N from raw indexes. All these value are reported in the result file.
If the macro reaches 500 dilation cycles without filling 99% of the image area, will report the result with a "higher than 500" (>500) notation. Lower percent indexes are not necessarily affected, and you can obtain H90% and H95% values lower than 500. However, a >500 result is not very informative and if you get a high number of those values you might begin to think about reducing image dimensions.
Please, note that the maximal number of dilation cycles and the 3 percent indexes showing in the results, are fixed parameters that can not be changed in this version of the macro. This might happen in future releases.