2001 Volume 40 Issue 1 Pages 30-39
A seamless three-dimensional (3D)Gamma-Compression Gamut Mapping Algorithm (GMA) is proposed. Currently Gamma-Compression GMA has been designed to work in two-dimensional (2D) Lightness-Chrominance planes segmented by primary and secondary color hues. The advanced 3D GMA is expected to work in 3D uniform color space without color segmentation.
This paper describes a 3D GMA based on the concept of Image-dependent. Considering the color gamuts' relations between source image and printing device, the Gamma-Compression GMA is applied in relation to the 3D shell shapes in CIEL*a*b* space. 3D shell shapes of source image and printer device are formed by connecting the most outside color points on their gamut surfaces. The surface polygon meshes or parametric cubic Spline surfaces are built up from these most outside color points, and true seamless 3D mapping is performed.
It is shown that the 3D GMA coupled with nonlinear compression and multi-mapping directions resulted in the better rendition than 2D GMA. Two kinds of methods, Image Lightness Division (ILD) and Focal Lightness Scaling (FLS), are introduced to find the optimum mapping directions.
