Hello
Have been working with PAD for some time now, and want to know a little about powerapps and AI.
If i have an AI model that can detect a road sign from a picture, how can I then use OCR on the detected part of the picture to translate it into text? (Is it possible)
Hi,
Here is what you could do
1. Get the bounding boxes of each detected object with Object detection
2. Call AI Builder Text Recognition
3. Get the bounding boxes of each text extracted
4. Use a condition to match the bounding boxes of the texts found with the bounding boxes of the object found.
E. g. Text."Coordinate top" > ObjectFound."Coordinate top"
AND Text."Coordinate left" > ObjectFound."Coordinate left"
AND Text."Coordinate top" + Text."Coordinate height" < ObjectFound."Coordinate top" + ObjectFound."Coordinate height"
AND Text."Coordinate left" + Text."Coordinate width" < ObjectFound."Coordinate left" + ObjectFound."Coordinate width"
Another option would be to crop the image to the detected object bounding boxes before trying to OCR it. But this requires to purchase some 3rd party connectors like Encodian.
Share
Report
1,187
All responses (
Answers (
