Connect HomeMarketplaceComponentsSkew Corrector using Machine Learning

Skew Corrector using Machine Learning

Machine Learning Model

Machine Learning Model

Updated: October 20th, 2020

Published: April 2nd, 2020

Downloads: 51

Industry: cross-industry

Community Support

Skew Corrector using Machine Learning

Machine Learning Model

Summary: 

The objective of this ML project is to correct the skewness in the image files using UiPath’s AI Fabric

The purpose of this project is to educate the end users on how to use UiPath’s MLSkill Activity with AI Fabric to correct the skewness in images. Those skew corrected images could be further used with Receipt Model, Abby or any 3rd party OCR for better extraction of data.
The file formats supported are PNG, JPEG or any scanned file in these 2 formats having textual data.
The project does not correct orientation.

Details

Benefits

During data extraction from a skewed document using OCR there is a high chance of losing some data or even extraction might fail.
Recommended way is to correct the skewness and those skew corrected images could be further used with Receipt Model, Abby or any 3rd party OCR for better extraction of data.

Limitations
Does not correct orientation

Compatibility

UiPath Studio >= 19.4
UiPath.MLServices.Activities

Dependencies

The input scanned files must have resolution equal and above 300 DPI (Dots Per Inch)
Internet connectivity of Bot to AI Fabric
UiPath Studio must have ML Skills Activity present (version 19.4 Onwards)
Download the package from uipath.mlservices.activities

Licensing

By clicking download you agree to the following license.