More from CGIV
More bits and pieces from CGIV today, typed in while watching England and Sweden battle it out on the telly... it's currently 1-1.
I saw a poster this morning describing how color morphological operations can be defined by defining a lexicographic ordering of two color pixels. I hadn't seen this before. I'd like to know if there is much application for this technique. F. Ortiz, "A New Method for Gaussian Noise Reduction in Colour Images by Colour Morphology."
Another interesting poster described an analysis of multispectral imagery of the Mona Lisa. The authors analyzed the effect of aging varnish and showed what the painting may have originally looked like. P. Cotte and D. Dupraz, "Multispectral Photography of the Famous Mona Lisa Painting."
A paper from the Unilever Food and Health Research Institute showed how to measure bulk rice color and other characteristics using a flat-bed scanner. G. van Dalen, "Colour Analysis of Rice Using Flatbed Scanning and Image Analysis."
Some of the lecture session presentations discussed content-based image retrieval. Researchers in this area must have a lot of patience and persistence. It's tough to even define measures of success.
One of the content-based image retrieval papers discussed using something called "local binary patterns." This looked like a fun idea for some MATLAB code and examples. Maybe I'll do a blog posting or two on it. D. Connah and G. Finlayson, "Using Local Binary Pattern Operators for Colour Constant Image Indexing."
A couple of presenters mentioned MATLAB. My favorite was the paper on a system for measuring the effectiveness of tooth whitening systems. This research was performed at the University of Leeds, with collaboration and sponsorship from Colgate-Palmolive. Presenter Wen Luo showed a screenshot of a MATLAB GUI used to facilitate the analysis. W. Luo, S. Westland, P. Brunton, R. Ellwood, I. Pretty, "Validation of a Tooth-Imaging System in Tooth-Whitening Trials."
Another presenter talked about a project to develop "film-like color rendering" in digital projection displays. Apparently, movie experts tend to be very biased one way or the other about film vs. digital, so the researchers had to come up with an innovative way to get their input on the system. They showed a frame projected using film, and then two different digital versions "A" and "B." The experts weren't asked to compare film to digital; instead they were asked to compare the two digital versions. J. Stauder, L. Blonde, J. Pines, P. Colantoni, A. Tremeau, "Film Look in Digital Post-Production."
My favorite phrases of the day:
- Earth movers distance measure
- Dead leaves noise model
- Golden eyes
Hold on ... GOAL!! England leads, 2-1. Beautiful header. What's the matter with me, I'm a baseball fan! Must be something in the water over here.
Hold on again, Sweden scores, now it's 2-2. And that's the way the game ends.
Comments
To leave a comment, please click here to sign in to your MathWorks Account or create a new one.