dsvg is an accessibility concept I've been stewing on for years, what if there was a way to give people more power to describe images by enabling the ability to describe specific sections?
dsvg leverages the built in features of svg that enable overlaying vector elements over raster graphics. This way in one file we can embed the image itself, the overlay vector graphics and utilize the vector graphics ability to embed data to create an interactive accessible experience with a fallback that screenreaders should be able to use
If you know me you know there's no monetary cost (and if you don't you know now). All of the code is under 100 lines and there's absolutely nothing that should be computationally expensive, though I will do further tests on worse devices to get a more solid picture at some point.
I'm still pretty sick at the moment so setting up a drawing tablet was out of the question, regardless I wanted to show more customized sections so I just drew them with my mouse. You can most certainly create better smoother sections.
I'm glad you asked! I do think I'd like to do a low end AR experiment with this where the descriptive tech can be overlaid on real images but as I'm working on other projects this will likely only materialize sooner rather than later if people express interest.