Use Google Vision API OCR engine

Use the Google Vision API OCR engine with IQ Bot to improve the accuracy of the optical character recognition (OCR) results for training documents in Asian languages, particularly in Japanese and Korean. Google Vision API OCR engine in IQ Bot supports all languages supported by the engine.

先决条件

Verify that the default port 443 is available to enable communication for the API endpoint using TCP.

注： See a list of allowed service URLs for using Google Vision API in the example section below.

Google Vision API files are installed on your machine automatically during the IQ Bot installation and is provided as a built-in plug-in, with a single-step installation, and is available for all users:

It is a cloud OCR engine.
Google Vision API does not support documents with more than one language. Before using this feature, ensure you want to primarily extract Japanese or Korean text only.
Languages that are written right to left, rather than left to right, are not supported within IQ Bot for document extraction. For example, Arabic, Aramaic, Azeri, Divehi, Fula, Hebrew, Kurdish, N'ko, Persian, Rohingya, Syriac, and Urdu.
Prior to this release, the engine primarily supported data extraction for Japanese and Korean text only.
Languages supported in Automation 360 IQ Bot
Data is not stored in Google Cloud and is removed immediately.
For additional information on privacy for data retention see: Google Data Usage

IQ Bot provides you with the license to use the Google Vision API OCR engine. You can also use your own license key.

Use your own license keys for Google Vision API OCR engine

注： Updates to Google Vision API such as OCR engine improvements, enhancements, and bug fixes might affect content extraction when you use this OCR engine in IQ Bot. Therefore, we recommend that you periodically refer to the Google Vision API release notes for the latest updates.

See Potential impact of Google Vision API OCR upgrade.

You have the following options for using this OCR engine:


Options	Description
Option 1: Set Google Vision API as your default engine	Configure the Settings.txt file. 限制： Configuring the settings.txt file is not applicable to Automation 360 IQ Bot Cloud
Option 2: Select Google Vision API engine directly from the UI when creating a learning instance	Select an OCR engine

过程

To set Google Vision API as your default OCR engine, configure the Settings.txt file as follows:

限制： Configuring the settings.txt file is not applicable to Automation 360 IQ Bot Cloud

Run the IQ Bot installer.
可选： Set Google Vision API as the default OCR for new learning instances.
Navigate to ..\Automation 360 IQ Bot\Configurations\Settings.txt, and modify the following value: OCREngine=GoogleVisionAPI.

This will set Google Vision API OCR engine as the default engine for your environment.
The GoogleOCREngineSettings.json file is generated within the ..\Automation 360 IQ Bot\Configurations folder with default values as follows:
Use the following with Google Vision API:
```
{
  "ADCJson": null,
  "EngineType": 2,
  "EnableFieldMerging": true,
  "EnableFieldRegions": true,
  "DebuggingFolder": "C:\\Users\\Public\\Documents\\Automation 360 IQBot Platform\\Logs\\Engine\\Google"
 }
```
For Japanese and Korean language documents, update the field: "EnableCustomParse": true.

注： A service restart is not required after making the change.

List of service URLs used for Google Vision API (optional)

Use the following allowed URLs if required:

注： All URLs point to a global service region in the US.

API URL: https://vision.googleapis.com/
Auth URL: https://accounts.google.com/o/oauth2/auth
Token URL: https://oauth2.googleapis.com/token
Auth Provider URL: https://www.googleapis.com/oauth2/v1/certs
Client Auth URL: https://www.googleapis.com/robot/v1/metadata/x509

后续步骤

Create a new learning instance with the Asian language document you want to train.
Train the document and set your learning instance to production.
Upload Asian language files and run the bot.
Download and view the accuracy of the extraction results.