I'm just writing to let everyone know that I just released a new experimental VOCR.
If you haven't heard VOCR, it's an app that lets you take screenshot of your mac screen, apply OCR, navigate the result, and move the mouse. It's similar to features on JAWS and NVDA.
MacOS Catalina finally included the OCR capability in their VisionKit that take advantage of machine learning.
In my opinion, it is much faster and accurate than Tesseract that previous VOCR utilized, and you don't have to deal with terminal and third party tools.
Here's an audio demo.
Because the tight security in Catalina, there are multiple steps for the initial setup. You can download VOCR and read the instruction below.
This is highly experimental, so use at own your risk!
Lastly, this kind of feature should be really built in to VoiceOver in my opinion. If you know anyone at Apple, let them know.