A hybrid intelligence approach to artifact recognition in digital publishing

J. Fernando Vega-Riveros; Hector J. Santos Villalobos

doi:10.1117/12.646240

10 February 2006 A hybrid intelligence approach to artifact recognition in digital publishing

J. Fernando Vega-Riveros, Hector J. Santos Villalobos

Proceedings Volume 6076, Digital Publishing; 60760B (2006) https://doi.org/10.1117/12.646240
Event: Electronic Imaging 2006, 2006, San Jose, California, United States

Abstract

The system presented integrates rule-based and case-based reasoning for artifact recognition in Digital Publishing. In Variable Data Printing (VDP) human proofing could result prohibitive since a job could contain millions of different instances that may contain two types of artifacts: 1) evident defects, like a text overflow or overlapping 2) style-dependent artifacts, subtle defects that show as inconsistencies with regard to the original job design. We designed a Knowledge-Based Artifact Recognition tool for document segmentation, layout understanding, artifact detection, and document design quality assessment. Document evaluation is constrained by reference to one instance of the VDP job proofed by a human expert against the remaining instances. Fundamental rules of document design are used in the rule-based component for document segmentation and layout understanding. Ambiguities in the design principles not covered by the rule-based system are analyzed by case-based reasoning, using the Nearest Neighbor Algorithm, where features from previous jobs are used to detect artifacts and inconsistencies within the document layout. We used a subset of XSL-FO and assembled a set of 44 document samples. The system detected all the job layout changes, while obtaining an overall average accuracy of 84.56%, with the highest accuracy of 92.82%, for overlapping and the lowest, 66.7%, for the lack-of-white-space.

Citation Download Citation

J. Fernando Vega-Riveros and Hector J. Santos Villalobos "A hybrid intelligence approach to artifact recognition in digital publishing", Proc. SPIE 6076, Digital Publishing, 60760B (10 February 2006); https://doi.org/10.1117/12.646240

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available