I am working to digitize these tables using google’s Form Parser, but have been struggling to get accurate replication. I’ve tried to read the table from an image into a csv, but it is still missing values, missing empty cells, or mixing values from different rows (even after preprocessing the image by increasing contrast, applying adaptive thresholding, and noise removal). Here is an example of the tables I’m trying to digitize: interpol crime statistics.

We’re using google OCR to read PDF or Images that are Loan Estimates.
We’re defining multiple fields such as loanTerm, loanPurpose
loanPurpose

We’re using google OCR to read PDF or Images that are Loan Estimates.
We’re defining multiple fields such as loanTerm, loanPurpose
loanPurpose

As part of our workflow we invoke DocumentAI ProcessDocument API (v1) API from our back end and the code has been in place for over 6 months and running without any errors. In the past one week we have been seeing random errors. But the main one being different variation of the error message “No remaining quota for ParseDocument”. When we check the ‘systems and quotas’ page we do not see anything which has even neared the quota limits and lost how to debug this further or resolve it. We make use of both OCR and Form Parser processors. Also, note that not all requests fail, only some of them fail.

I can see how many testing documents the model was evaluated against, but can’t see how many documents where used to train the model. Do I just assume that it’s all documents, is there a way to view this?

I’ve trained a custom form model using Google Document AI Custom Extractor. According to the official Document AI documentation and the Google Cloud platform interface, it seems that the only way to use this model is through the API.