# How much accurate is OCR

Using flexicapture its goood too use.

##### Post Metadata
- Posted at: over 6 years ago
- Author title: Seeking Learning in the field of digital transformation along with the knowledge of new technologies
- Net upvotes: 1


## Comments
### Comment 1

The accuracy of OCR depends on the document quality.
For unstructured documents embedded OCR tools (from Google, Microsoft) are fairly acceptable. 
For semistructured documents (invoices &amp; receipts &amp; bank statements) we are working with Abbyy FlexiCapture tools. 
For each document type it is possible to define a layout for data that must be extracted using Abbyy FlexiLayout. This layout is then used in Abbyy FC Document Definition to test the extraction. This will generate so called .fcdot that can be integrated in UIPath workflows using Abbyy Intelligent OCR package. In this way the extraction is very accurate and automation works excellent. Still, the confidence level should be defined in Abbyy FC.  In the case of any uncertainties, UIPath raises Abbyy Verification station. It is possible to extract data not only from pdf files, but also from .tiff, .png and .jpg. 
Still, to use Abbyy Intelligent OCR package is mandatory to use Abbyy FC Engine. This is licensed separately by UIPath directly and not by Abbyy.
One last recommendation: 
Use Abbyy FC Standalone instead of Abbyy FC Distributed. It is mandatory to use integration in UIPath activities for the .fcdot files. 


##### Comment Metadata
- Posted at: over 6 years ago
- Author title: Head of RPA Services



### Comment 2

The accuracy of OCR depends on the document quality.
For unstructured documents embedded OCR tools (from Google, Microsoft) are fairly acceptable. 
For semistructured documents (invoices &amp; receipts &amp; bank statements) we are working with Abbyy FlexiCapture tools. 
For each document type it is possible to define a layout for data that must be extracted using Abbyy FlexiLayout. This layout is then used in Abbyy FC Document Definition to test the extraction. This will generate so called .fcdot that can be integrated in UIPath workflows using Abbyy Intelligent OCR package. In this way the extraction is very accurate and automation works excellent. Still, the confidence level should be defined in Abbyy FC.  In the case of any uncertainties, UIPath raises Abbyy Verification station. It is possible to extract data not only from pdf files, but also from .tiff, .png and .jpg. 
Still, to use Abbyy Intelligent OCR package is mandatory to use Abbyy FC Engine. This is licensed separately by UIPath directly and not by Abbyy.
One last recommendation: 
Use Abbyy FC Standalone instead of Abbyy FC Distributed. It is mandatory to use integration in UIPath activities for the .fcdot files. 


##### Comment Metadata
- Posted at: over 6 years ago
- Author title: Head of RPA Services



### Comment 3

Most of the times OCR is fairly accurate. But when there are similar texts on the screen it behaves ambiguously. For instance if you have  &quot;mode-I&quot; and &quot;mode-II&quot; on the same page it might confuse between the two.

##### Comment Metadata
- Posted at: over 6 years ago
- Author title: Graduate Engineering Trainee at Vodafone Shared Services India



### Comment 4

100% accurate in especially for PDF it gonna use full a lot

##### Comment Metadata
- Posted at: over 6 years ago
- Author title: RPA Sr. Consultant at Finonyx Software Solutions Pvt Ltd



### Comment 5

very accurate with the last release of uipath screen ocr is free and better than abbyy cloud or google cloud even faster

##### Comment Metadata
- Posted at: over 6 years ago
- Author title: RPA Consultant en TEAM International



### Comment 6

It depend on the content you are scanning. my assumption is if content is clear then it can give 100% but if content is blur or hand written then it can reduce to 50% depend on the clarity of the content

##### Comment Metadata
- Posted at: over 6 years ago
- Author title: Tech Lead



### Comment 7

UiPath uses many different methods of text recognition and reading, including multiple OCR engines such as Microsoft and Google, all of which are very reliable.

##### Comment Metadata
- Posted at: over 6 years ago
- Author title: Intelligent Automation Specialist | RPA, AI and Digital Solutions | Hackett Consulting Services
#HCS #RiseOfTheMidMarket




## Related Product
[UiPath Agentic Automation](https://www.g2.com/products/uipath-agentic-automation/reviews)

## Related Category
[Robotic Process Automation (RPA)](https://www.g2.com/categories/robotic-process-automation-rpa)

## Related discussions
- [How well does Trello scale into a larger team?](https://www.g2.com/discussions/1-how-well-does-trello-scale-into-a-larger-team)
  - Posted at: about 13 years ago
  - Comments: 6
- [Can we please add a new section](https://www.g2.com/discussions/2-can-we-please-add-a-new-section)
  - Posted at: about 13 years ago
  - Comments: 0
- [Quantifiable benefits from implementing your CRM](https://www.g2.com/discussions/quantifiable-benefits-from-implementing-your-crm)
  - Posted at: about 13 years ago
  - Comments: 4


