This repository contains a mod of the standard ClickCoordinatesTool macro from ImageJ, better suited for our image annotation requirements.

ImageJ is a popular open source image processing program designed for scientific multidimensional images. It already includes a macro for image annotation, which does not scale well for bigger pictures and crowd-counting. We slightly modified it with this very simple snippet:

  setLineWidth(3);
  setColor(0,255,0);
  tickLength = 9;

so that green crosses are drawn for each click and they are bigger than the default size (otherwise the crosses were undetectable on 16+ MP photos).

This macro requires a valid installation of an ImageJ distribution. There are many of them, including our preferred, which is called Fiji (whose meaning is a recursive acronym: Fiji is just imagej). Download it from its website. In case you choose a non-JRE version, you have to provide a valid java installation to the software.

Many of these distributions already include a basic ClickCoordinatesTool.txt macro, which is not suitable for our requirements! So please use our tool instead, downloading this file to your computer: ClickCoordinatesTool.txt (right click on the link and select Save link as...)

To annotate a picture follow this procedure:

1. Open your ImageJ software
2. Click on File, then Open and select an image on your computer
3. Click on Plugins, then Macros, then Install... and then double click on the file ClickCoordinatesTool.txt (our version!) which you downloaded previously
4. Click on each and every person you can distinguish in the picture. This is the most important step and you should devote enough time to it. Do not click on an item if you are not sure it is a person, and please put the crossmark precisely on the people's heads.
5. In case you need to zoom in/out of the picture, please use the + and - keys on your keyboard.
6. In case you put a cross by mistake, please follow this other step accurately: go to Edit and then Undo into the image window and inside the Results window select the last entry and remove it going to Edit and then Cut.
7. At the end of the annotation, save the picture going to File, Save as, Jpeg... and then giving it a name like the original but with _ann appended to it. You also have to save the Results, going to the Results window and saving it through File, Save as and then giving it a name which is the original name but with _count appended to it. Example: from a given image.jpg you should obtain image_ann.jpg and image_count.xls