I believe the v4 APIs are better at this for bounding boxes, though I'm not sure if they expose anything with true semantic segmentation at the pixel level yet.
I don't think Smart Cropping is right for this as its largely autonomous, however, Form Recognizer would be able to handle reading the text on the check.
@@cissemy Yep. I got it. In that case I'd nudge you towards Google Document scan. It's less automated than you could get with Azure Cognitive Services, but it's built to digitize images of documents for storage later on.
Code available on GitHub at github.com/IntegerMan/AutomatingMyDog
Thanks,
nice tool, do you know if Azure have something about image-segmentation ?
I believe the v4 APIs are better at this for bounding boxes, though I'm not sure if they expose anything with true semantic segmentation at the pixel level yet.
Thanks
Let say I have a check on my desktop .If I take the picture of the check with the whole desktop , Can azure keep only the check part ?
I don't think Smart Cropping is right for this as its largely autonomous, however, Form Recognizer would be able to handle reading the text on the check.
@@MattEland
I want to keep the picture of the check only and remove all the surroundings.
Thanks
@@cissemy Yep. I got it. In that case I'd nudge you towards Google Document scan. It's less automated than you could get with Azure Cognitive Services, but it's built to digitize images of documents for storage later on.
@@MattEland
Does Google Doc have any c# code for that ?
Thanks
I think it is just like canny edge detection