The real "brains" behind the standard in the company I was in at that time (KTAS) which was a telephone company.....was Birger Niss and Jørgen Vaaben:
https://jpeg.org/items/20170813_press.htmlIt seems JPEG was recognized as a standard first in 1992 but it was developed some years before that in a "ESPRIT" project.
Telecommunication lines were slow at that time (ISDN = 64kbit/s). The goal was to be able to search in picture databases fetching pictures at that speed and my goal was to write a JPEG decoder that could do it in realtime on a IBM XT machine or something similar. Therefor the JPEG standard has an option called progressive mode where you first get a very rough image and then it gradually gets finer and finer so you fast can decide if it is the wanted image or you want to look at the next.
The JPEG standard works on 8x8 pixel blocks (run-length coding). So it is one long bit-stream like if you receive a morse coding message. In progressive mode first scan has equal 8x8 pixels and then it gets finer and finer. A picture is coded as luminance + two chrominance components. The luminance component has nearly all the information. I was surprised that the chrominance components that so little information was needed.......when I remember back :-)
So it was an advantage to go from RGB to Luminance + Chrominance. The image could be better compressed.
Think this explains it quite good (I had to go deep in my memory):
https://yasoob.me/posts/understanding-and-writing-jpeg-decoder-in-python/ Yes, we could meet in Copenhagen one day and do some tests.......
