Closed duncdrum closed 3 years ago
In GitLab by @JoernT on Nov 6, 2018, 17:16
@adober yes, this should in principle be possible. However i cannot tell how much effort it takes to extract the right part of the image as the selection is on another canvas and selection coordinates would need to be mapped to actual image canvas positions to extract the part of the image you want to OCR.
In GitLab by @MatthiasArnold on Apr 30, 2019, 18:58
Hm. The feature agnes asks for is not of high priority.
however, i do not really understand this part of @JoernT 's answer:
the selection is on another canvas and selection coordinates would need to be mapped to actual image canvas positions
the reason why we are building the annotation app is exactly to be able to further process the annotated parts of the image. so if that is a problem, the whole thing somehow missed the target... but i assume i'm just misunderstanding something? ;)
In GitLab by @duncdrum on Jun 28, 2019, 12:47
Yes so this seems to have increased in the priority list. I m curious as well. @JoernT Where are those coordinates pointing to?
In GitLab by @JoernT on Jun 28, 2019, 15:59
@duncdrum all shapes are svg polygons that are stored in the annotation. These are relative to the canvas we're drawing on. Actually there are 2 canvases overlaying each other and having the same size. The OpenSeaDragon canvas and a Fabricjs canvas. I'm happy to explain in more detail in the code next week as it would be hard to give all this information here.
Please re-open if original question is still unanswered.
The original issue
could not be created. This is a dummy issue, replacing the original one. It contains everything but the original issue description. In case the gitlab repository is still existing, visit the following link to show the original issue:
TODO