Thanks to all who attended the digital images workshop at Birkbeck in London on 14 February 2017. It was great to test out the manual and batch processing functions of ImageJ, and there seemed to be a number of applications for the measurements that can be done using the software, sometimes involving the angle of lines set on the image, e.g. the direction of eyes in a painting or photograph, sometimes about the calculation of area, such as the treatment of lighter, foregrounded details against dark backgrounds.
Whereas the manual tools seemed quite intuitive, the challenge arises with the use of macros and batch processing. We worked with Lev Manovich and the Software Studies Initiative’s ImagePlot package, which contains a number of macros allowing users to extract features, such as median brightness, from a large set of image files. It also allows you to create scatterplots of those data points and overlay thumbnails of the images. At least in theory!
We encountered a few errors when trying to start the plots, but after downloading the ImageJ (64-bit) package from the National Institutes of Health website, the data seemed to work better with the macros on the Mac OS, and we managed to plot the visualisations.
Another point to remember when using ImageJ is to have Java installed: https://www.java.com/en/download/manual.jsp – this slowed us down a bit at the beginning of the day, but we got the software up and running.
We spent less time on our other software, ANVIL and Cinemetrics, but I am considering how the former might be intergrated into our workshops on language data analysis at the University of East Anglia. The coding specifications are absolutely fascinating for research into the interaction between gesture and language.