My ongoing doctoral research into multimodal assessment includes diverse data, of which a substantial amount is visual. That is, the data includes still or moving images taking into account colours, design elements as well as embodied modes (such as gaze and gestures).
If extracts from this kind of data should be included in any form of research publication (of course with consent from participants) images need to be processed to maintain the confidentiality of research participant. If people are present in an image that should be used for evidencing or illustrating some argument in the research, this person or these people need to be represented in a way that does not reveal their identity but still visualises the example or argument being made.
With the increasing use and capabilities of AI how we use and publish images need careful consideration. To reduce the risk of data being extracted from images I am now ‘doctoring’ images and illustrations using offline programs. Depending on what feature I want to highlight in an image I use different strategies. Word comes with a set of cartoon characters (full, half body, heads, faces) that can be superimposed on a photo for example. I then take a screenshot of this ‘new’ image’ where the person is now represented by a drawn character. By taking a screenshots I avoid metadata that is present in the original images (even if this has been manipulated). I some cases I want to keep the visual information from gestures, body language and placement in a room. Then cartoon elements to mask a person in my image are not a good choice, as the image would loose its relevance as an illustration of for example embodied modes. One solution is to manually draw from a photo, only keeping the relevant information. If I instead chose to use a digital method I now work with images in Inkscape. Sometimes in combination with cartoon elements in Word.
Below, I have used an old family photo (ca 1930), and processed it in Inkscape. The original is top left. The second image has been sharpened, the third is a filter that turns the photo into a line drawing and the fourth (top right) is another variation of line/ink drawing from photo. The bottom row of photos are examples with more or less use of further masking.
Original photo (by unknown) ca 1930, Boy with dachshund.
I have not spent a lot of time on processing these images, but they should offer some ideas of what you can do with just a bit of effort. There is much more you can do if your put your mind to it. It all depends on the image, your purpose and the level of integrity needed. In some situations the solution would be to not use images at all but to describe instead.
However, if you do need to work with images then Inkscape is a free software. It can be a bit tricky to start with. On the other hand there are lots of ‘how to do’ videos and instructions for Inkscape online. Easier to start with are using cartoon people in Word.
multimodal study 