TABLE OF CONTENT
1. Introduction to Azure Form Recognizer2. Prerequisites3. Step by Step Guide Conclusion 5. CloudThat 6. FAQs1. Introduction to Azure Form Recognizer
Azure Form Recognizer, a cloud-based Azure Applied AI Service, extracts key-value pairs and text from your documents using AI powered data solutions. Form Recognizer scans your documents and forms and extracts text and data. It maps field relationships to key-value pairs and returns a structured JSON output.
This blog is about Automate Report Extraction using Microsoft Azure Cognitive Service: Part 1 which can help you refresh and revise your understanding.
Let’s get started with the Form Recognizer sample labeling tool:
The Form Recognizer Sample labeling tool is an open source tool that allows you to test the latest features in Azure Form Recognizer as well as Optical Character Recognition (OCR).
Layout API allows you to analyze documents: Extract text, tables and selection marks from documents.
Prebuilt models are useful for analysing documents: Use a prebuilt model to extract data out of receipts, business cards, and identity documents.
Analyze and train a custom form: A custom model can be used to extract data from documents that are specific to business data and use cases.
2. Prerequisites
To get started, you will need the following:
An Azure subscription
A Cognitive Services or Form Recognizer resource. Once you have an Azure subscription, you can create a single-service and multi-service Form Recognizer resource on the Azure portal to obtain your Key and endpoint.
3. Step by Step Guide
Step 1: Sign in at the Azure Portal, search the search bar for form recognizer and then select the service
Step 2: Click Create form recognizer.
Step 3: Fill in the details below.
Select the right subscription
If you have an existing Resource group, create a new one or select the desired Resource group.
Select the region in which you would like to deploy the service
Please provide the correct name for the service
Click on Review + Create to select the F0 pricing tier, which is the free tier.
Step 4: After the service has been deployed, take a note of the Keys and Endpoints that will be needed in the next steps
Step 5: Create a storage account with the correct name, region, as well as other details.
Click on Review + Create
Step 6: Select the containers in the left panel and create another container to store images. The public access level should not be Blob
Step 7: Click Upload. Upload the images you wish to use for the custom model of the form recognitionr
Step 8: Select the storage accounts created in Step 5 and select the CORS from left panel. Configure the settings as shown below, and click on Save
Step 9: Create the SAS URL, as shown below
Step 10: Copy the Blob SAS URL and add the container name to the URL, as shown below.
Step 11: Go to https://fott.azurewebsites.net/ and click on the Connection symbol in the left panel
To add a new connection, click on the “+”.
Step 9: Name the connection and the SAS URL.
Click on Save Connection
Step 12: Once you have established the connection, go to Home, select your custom model, and click on New Project
Step 13: Fill in the details below.
Please give the project name
Select the connection that was created in Step 11.
In Step 4, indicate the endpoint of form recognizer.
Enter the Key of the form recognitionr in Step 4 and click on Save Project
Step 14: Once you have created your project, the images that were stored in the container can be loaded onto the left panel
Click on “+”. to add the tags shown below. Let’s say you need to extract the Name and No. The number of days, and the status of the leave taken from the custom images
Step 15: Select the data from the image, and then select the tags. All images should be tagged as shown below
St