“Optimized color LUT transformations by means of analysis of image memorable and subject important colors” by Gurieva, Guryev, Puente and Lopez

  • ©Natalia Gurieva, Igor Guryev, Francisco Javier Montecillo Puente, and Reynaldo Thompson Lopez

  • ©Natalia Gurieva, Igor Guryev, Francisco Javier Montecillo Puente, and Reynaldo Thompson Lopez

  • ©Natalia Gurieva, Igor Guryev, Francisco Javier Montecillo Puente, and Reynaldo Thompson Lopez

Conference:


Type:


Title:

    Optimized color LUT transformations by means of analysis of image memorable and subject important colors

Presenter(s)/Author(s):



Abstract:


    In this work we are investigating ICC color management workflow in color imaging system and proposing improvement to standard ICC-based color management scheme. We are created optimal look-up tables of ICC profiles by using spline approximation and neural network for direct transformation RGB → Lab and regression models for inverse transformation Lab → RGB and integrated criterion (optimization criterion). Our algorithm implements corresponding color reproduction to accurately predict memorable and scene-important colors. Mean errors was reduced by 21% as compared to standard color profiles for particular output device and type of paper.

References:


    1. Phil Green (Ed.), Michael Kriss. Color Management: Understanding and Using ICC Profiles, Wiley-IS&T Series in Imaging Science and Technology, 2010, 314p.
    2. Specification ICC.1:2010 (Profile version 4.3.0.0) Image technology colour management — Architecture, profile format, and data structure. International Color Consortium. http://www.color.org/specification/ICC1v43_2010-12.pdf


ACM Digital Library Publication:



Overview Page: