Use Microsoft Azure Computer Vision OCR engine

Microsoft Azure Computer Vision OCR engine is a cloud OCR engine and is available to all customers of IQ Bot. IQ Bot automates some installation steps for Microsoft Azure Computer Vision OCR engine.

When you process documents using the Microsoft Azure Computer Vision OCR engine, data is removed immediately. Microsoft Azure Computer Vision OCR engine provides approximately 18% STP and 80% accuracy with data extraction. For more information on privacy for data retention, see Microsoft Azure Cognitive Services.
Note: IQ Bot also supports containerized deployment of Microsoft Azure Computer Vision OCR engine:

Configure IQ Bot using Docker container and Azure OCR (A-People login required)

When creating a learning instance, you can select any language from the Primary language of documents drop-down menu. During processing, the OCR engine tries to auto-detect the primary language and can override your selection. For example, you can specify English, but if you upload a Spanish document, the API will try to process the Spanish language.

Tip: You can troubleshoot any issues with extracting data from low quality or handwritten documents IQ Bot unable to extract data from low quality and Handwritten documents (A-People login required)
Follow the steps to configure the Settings.txt file and use Microsoft Azure Computer Vision OCR engine.
Important: Configuring the settings.txt file is not applicable to Automation 360 IQ Bot Cloud

Procedure

  1. Ensure your IQ Bot server has internet connectivity and external DNS resolution.
    Check to enable communication for the API endpoint on default port = 443: default endpoint = https://aai-iq-bot-ocr.cognitiveservices.azure.com/.
    Note: To validate connectivity to the API endpoint and the server, open a web browser on the server and type in the complete URL for the API endpoint. If the browser returns a 404 error, then it implies that a response from the API endpoint was received.
  2. You can use your own Microsoft Azure Computer Vision OCR engine subscription or spell-check keys. See Use your own keys for Microsoft Azure Computer Vision OCR engine for more information.
  3. Optional: From the C:\Program Files (x86)\Automation 360 IQ Bot <version number>\Configurations folder, open the Settings.txt file, and change the OCR engine value to OCREngine=Tesseract4 or OCREngine=Abbyy to OCREngine=MicrosoftAzureAPI. Then save the file.
  4. Run the stopanduninstallallservices.bat file at C:\Program Files (x86)\Automation 360 IQ Bot\Configurations.
    After stopping the services, run installandstartallservices.bat file to install and start the services.
  5. Create learning instances in IQ Bot, and use Microsoft Azure Computer Vision OCR engine for the text segmentation, and OCR engine for these learning instances.
    For these learning instances, you can continue to use the IQ Bot capabilities for document classification, auto-mapped fields, cognitive extraction, and field value auto-correction.
List of service URLs used for Microsoft Azure Computer Vision OCR engine (optional)
  • https://aai-iq-bot-ocr.cognitiveservices.azure.com/
  • https://aai-iq-bot-ocr-spellcheck.cognitiveservices.azure.com/
Note: All URLs point to a global service region in United States.

If IQ Bot is unable to process a document using Microsoft Azure Computer Vision OCR engine, you can review the troubleshooting tips.

IQ Bot unable to process the document with Microsoft Azure OCR (A-People login required)