This post is about base64 images and how their resurgence may affect the usefulness of Google search results.
Google does not index base64 images (the image in this post is base64 for example).
However, Google itself uses base64 to speed up the web. In fact if you are using Google page speed service, your images may well be out of the Google search results already.
Base64 images are images that are represented by code (letters, numbers and characters) rather than by a file. Just about any image anywhere can be turned into base64 by using online base64 image converters.
Why are people using base64 more now?
To make their pages load faster. Pagespeed is probably one of the things that most webmasters can agree on. Faster is better. If you were to have a webpage that had text and one image, that page would normally take at least two round trips to a webserver to display. If you used a base64 image however, it would only take one. Because the image would be part of the html file rather than an additional thing to download.
Why isn’t everyone doing this then?
The main reason is likely because Google does not index images that use base64 encoding. Unfortunately this means that webmasters have to make a choice between serving their users the best way they can, or serving Google.
What happens if you use base64 images?
Currently it means Google will not index your images and they will not be seen in image search or other such Google entities.
This is bad for education
I have a website (feedthebot.com) that is used by hundreds of thousands of people every month and virtually every image on that site is base64 encoded. I do this fully knowing that those images will not be indexed by Google.
This might seem strange to most people, but it is natural choice to me.
I make things for my users. I do not make them for Google.
So if more and more responsible authors start using base64 image to serve their users better, then Google image search will become less and less accurate or useful.