OCR function not reading

[Craig_Clancy1]

Just checking if there are any issues with the OCR models. I have trained the app and the patterns are all as I want them. The images show the respective column areas correctly, however when I use the app and take an image the data is not populating into the related fields.
Am I missing a step somewhere?
Many Tx for any guidance.

[Koichi_Tsuji]

Not reading at all, or it is filling to some fields, but not for some other fields?

[Craig_Clancy1]

Hi and thanks for responding. No fields are filling at all.

[Koichi_Tsuji]

Sure.
I would say OCR model is bit buggy , Appsheet team should be working on it for improvement, also with additional new features.

@tony

[Craig_Clancy1]

I sent a request to support yesterday so hoping to hear back soon.
Its been sitting in Beta for a long time now having had plenty of intetest initially.

Wonder why no solid movement to full priduction as it is a very handy feature.

Thanks for your help.

[Koichi_Tsuji]

Sure, thank you. Will see.

[adithya5498]

@tsuji_koichi Thanks for your support. I also faced some issues!! I was trying to train these below pictures, But it is not recognizing as N N N nor Y instead of this it is recognizing as ZZZZZ sometimes YZZZ below I have attached the OCR extracted text using the expression OCRTEXT([Image Column]). Please help me on this Thanks in Advance!

[Koichi_Tsuji]

This is simply my assumption, but Appsheet OCR engine (it should be Googel Vision API) is reading text vertically for some part of your image.
We write English always from left to right horizontally, but my own language (Japanese) we write from left to right vertically, but also we can right top to bottom horizontally.

Out of curiosity, I just read image where the characters line is in horizontal order, but we managed to read the text correctly. (impressed)

@Takuya_Miyai

@Kirk_Masden