Is there a way to map the page number along with the tables, as we have to extract from a table which spans across multiple pages?
Is there a way to map the page number along with the tables, as we have to extract from a table which spans across multiple pages?
Hi, thanks for reaching out. The Layout API has "pages" parameter that enables you to specify pages for multi-page documents. The results would include information such as page number, page results, number of tables on each page. Hope this helps, otherwise, please elaborate a little more so we can provide better guidance. Thanks.
Basically i am trying to use OCR form tools to train the model for a multi page document for which the table data spans across multiple pages. How am doing is, am marking tables in each page(in my case, 5 page document) as a separate table(table 1, 2 3 ,4 and 5 ) and training it.
while training the model, am getting the problem as some of the tables in certain pages for which the confidence score is just 0%. the system is not recognizing those tables on certain pages.
When i try to analyze the document with the model been trained, respective page table data has been retrieved only for first 2 pages and rest of the tables modelled for Page3, 4 and 5, its trying to pull the data from the table in wrong page(Page 2 in my case)
Am using Custom Form - Analyze Form, API to extract the data.
It has a variable number of rows. We have used Configure table tables tags. I could not able to navigate to the link which you have mentioned above.
4 people are following this question.
Azure Cognitive Services (Form Recogniser) keeps timing-out for the past two days
Trying to open a file and it says select the encoding that makes your document readable
Integrating Constant Contact with D365 Marketing Leads/Activities
Form Recognizer Sampling Tool unable to access files in Container