I came across the work of the artist Petros Vrellis recently and was fascinated how a technique we use in CT image reconstruction is being used to create art.
The artist weaves a continuous line of thead in straight lines between nails around the periphery of an aluminium frame to build up an image. The more threads between the pegs the denser the line.
The technique is called threading or weaving and is similar to how back projection is used in creating CT images.
The clever bods out there have used his algorithm to enable us to reconstruct any picture. (I suspect the resultant image may benefit from a reconstruction Kernel as used in CT to reduce noise and sharpen the edges - but this art not science!)
If you would like to have a play you can upload an image and try altering the number of nails and threads used as well as the thread thickness etc.
Here is one I made earlier using my profile photo from this forum.
Hereâs one I did of my kids using 200 pins and 15,000 coloured threads. Interesting to see the concentric circle artefact we had with early CT. I canât recall the name of that? Was it Moire?
I think itâs because the thead spacing defines your sampling freqency. Notice how the artifacts result from different cross-ver densities of your âraysâ. The rings in the middle will be due to the pixel sampling of your image of the image I think and how that corresponds with your thread-made âpixelsâ. In which case it is indeed Moire. When you look in real life do you see the circle artifacts or are they only there in your picture of the picture? Also, are the treads for the image generted from a digital image (I am pretty sure yes) - what is the pixel resolution of that image compared to the thread spacings? The thread spacing is not evenly disctibuted with respect compared to image pixel spacing ⌠hence the variation in the artifacts across the image. Very nice - thank for posting
