Download presentation
Presentation is loading. Please wait.
1
3.3 Screening Part 3
2
Start With a Lattice Continuous Parameter Halftone Cell (CPHC)
3
Finding Discrete Parameter Halftone Cell (DPHC)
Compute number of pixels in unit cell = |det(N)| Assign pixels to unit cell in order of decreasing area of overlap with CPHC Skip over pixels that are congruent to a pixel that has already been assigned to DPHC DPHC Area
4
Threshold Assignment by Growing Dots and Holes Simultaneously
Abs. = 0.74 Abs. = 0.26 Abs. = 0.53
5
Representing Non-rectangular Halftone Cells by Rectangular Regions [Holladay, 1980]
Parameters
6
Perceptual model: color device Neugebauer Primaries Ri()
CIE XYZ CMF’s Neugebauer Primaries Ri() D65
7
Opponent Color Channels
Perceptual Model (Cont.) Opponent Color Channels Use linearized version of L*a* b* color space to represent opponent color channels of the human visual system Flohr et al [1993]
8
Spatial Frequency Response of Opponent Channels
Perceptual Model Spatial Frequency Response of Opponent Channels Luminance [Nasanen] Chrominance [Kolpatzik and Bouman] cycles/sample cycles/sample cycles/sample cycles/sample Lin and Allebach [1998] used similar model to design color FM screens
9
Overall Framework for Perceptual Model Part I
10
Overall Framework for Perceptual Model Part II
11
Error Metrics MSE/pixel at jth Level Two Metrics
12
Constraints for Optimization
Search Strategy Optimize for textures along the neutral axis Exhaustive search with constraints on the following: Density (area) of periodicity matrix Screen angles Vector lengths in periodicity matrix Constraints for Optimization Same Density (area): Screen angles: Vector lengths in periodicity matrix: Vary 3 variables of N and compute 4th to satisfy fixed density (area) Set cyan offset to be the null vector
13
Experimental Results Parameters Typical Angles Density (area) = 16
Vary elements of from -3 to 3 Vary elements of from 0 to 3 Typical Angles Screen angles separation = 30o Yule [1967] Cyan = 0o, Magenta = 33.69o, Yellow = 68.19o
14
Magnified Scanned Textures for Various Screens
Absorptance = 0.25 Worst Best MSE = 9 x Best Optimized for Registration Errors Typical MSE = 4 x Best MSE = 5 x Best
15
Magnified Scanned Textures for Various Screens
Worst Best Absorptance = 0.5 MSE = 16 x Best Optimized for Registration Errors Typical MSE = 4 x Best MSE = 6 x Best
16
Weighted Spectra of Error in YyCxCz
Best Worst
17
Weighted Spectra of Error in YyCxCz
Optimized for Registration Errors Typical
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.