A non-sensical experiment at using the comment field in an images EXIF info to store pixels hidden by a baked-on watermark.
A section of the image is analyzed, and the pixels are transformed and stored in the comments field of the exif info.
Then, a watermark is applied to the same spot of the image.
On the client, the image is loaded in HTML and the exif is read. The un-watermarked pixels are draw to a canvas element which overlays the image.
It works, but it's kind of useless.
Presented here for educational purposes, and because I want to free up a hidden repo.