absent-sapphire•2mo ago
OCR, not automatically using the image as the example image.
I don’t know if it’s an already available function and I just didn’t figure it out, but it would be much easier if once I took a picture/shared an image to migaku app’s OCR, the app would then use said image as the example image in each card created from the mined text! Also is it still not possible to create a card with all the unknown words in the mined phrase? (Only way I found to do this in the app is just creating a card from one unknown term and then manually modifying the card).
13 Replies
When you use Migaku OCR tool it does already use the image in the flash card you make
absent-sapphireOP•2mo ago
No it doesn’t, the app produces an AI generated image instead of using the original one from which the text was extracted.
They don’t generate AI images, they pull from bing. I was also just making OCR flashcards this morning and it used the image. Could you send a video of what’s happening?
absent-sapphireOP•2mo ago
My bad, that explains why they are usually so out of context… anyway here is a video
ah yeah, if you send it to the clipboard, for some reason it wont save the photo you were using, i have no clue why they did this since its just such an odd and annoying decision imo... unless this is a bug but i dont think it is? @HulK
absent-sapphireOP•2mo ago
I finally get it now… but yeah… trying to select the right word from the mined text before sending it to clipboard is such a pain I basically stopped doing it from that screen. Hope this changes… thank you for your help!
optimistic-gold•2mo ago
What's painful about selecting it from the other (non-clipboard) screen?
absent-sapphireOP•2mo ago
When trying to “tap” a word on that screen the app often selects the wrong word. Also it sometimes crashes. Moreover the phrase is not editable on that screen so if the OCR has a small mistake on the word I want to select I have no way of fixing it, like I can instead do on the clipboard screen.
optimistic-gold•2mo ago
Well for now you can hit
Add in the Card Creator and find the image that you OCR'd
@The Coding Ape Do you know why selection would be less reliable on OCR screen as opposed to Clipboard?deep-jade•2mo ago
Can you explain tap on screen? Do you mean text on the image or text on bottom sheet?
absent-sapphireOP•2mo ago
To select the text on the image I always use “select all” as it’s way easier. I’m referring to tapping on a particular word on the bottom sheet after the text has been extracted. The apps often seems to not recognize the “limits” of each word as well as when the text is sent to clipboard, so often tapping on a term results in the app opening either a completely different word in the phrase or a purely parsed word. Also if the OCR has made a mistake in extracting the text I cannot manually fix it and so selecting a term becomes impossible. I don’t know how to be more clear in my explanation sorry…
deep-jade•2mo ago
Ok. i got your point. but yeah, that part of text selecting is from iOS default API 😦
absent-sapphireOP•2mo ago
Got it, then I’ll just do as said above and manually add the image I want each time. Thanks