Oil Painting Classification

By Shiyu Luo
Dec. 2010
 Motivation and Goal
 Methods
 Feature extractions
 Classification Results
 Analysis and conclusion
 References
Motivation and Goal
 Oil paintings are of great value
 Art
 History
 Even more counterfeits make it harder to identify the
authentic works
 Traditional: signatures, Dates and producers of
canvas, etc.
 Proposal: by Digital Image Processing
Brushwork example of one of da Vinci’s painting
Left: Brushwork in original painting
Right: micro-view of grey-degree of the red square
 In this pilot project, painting-based approaches are
 Data set: 8 X-rayed paintings from Leonardo da Vinci
 Method:
 Patch selection
 Feature extraction
 Multi Layer Perceptron
Feature extraction
 General requirements:
 Intra-class variance must be small
 Inter-class separation should be large
 Independent of the size, orientation, and location of the
 Four features are employed
 Fourier Transform (Brushworks)
 Wavelet Transform (lower resolution image)
 Statistical Approach (texture)
E.g., 2nd moment: a measure of gray-level contrast to describe
relative smoothness
 Covariance Matrix
Multi Layer Perceptron (MLP)
 MLP: Error Back Propagation
A diagram demonstration of Multi Layer Perceptron
Analysis & Conclusion
 Generally speaking, C_rate can be achieved at
around 40% - 50%
 50x50 patch-based generally achieves better and
more stable results than 100x100 patch-based does.
 For 50x50 patch-based, the better and relatively
stable results are those with 6-8 neurons in hidden
 Those “excellent” results of 100x100 maybe I’m
“luck” in the 3 trails.
Future work and improvement
 X-rays maybe one of the limits on achieving better
classification rates; colored paintings could be used
in the future
 2nd or higher order wavelet transforms maybe used
to improve the feature vector
 Other neuron network methods are to be tested to
better suit this painting classification problem
Selected References
 Siwei Lyn, Daniel Rockmore, and Hany Farid. A digital technique for
art authentication. 17006-17010, PNAS, Dec. 2004, vol. 101, no.49.
C. Richard Johnson, Jr., Ella Hendriks, Igor J. Berezhnoy, Eugene
Brevdo, Shannon M. Hughes, Ingrid Daubechies, Jia Li, Eric Postma,
and James Z. Wang. Image Processing for Artist Identification:
Computerized Analysis of Vincent van Gogh’s Painting Brushstrokes.
Jana Zujovic, Scott Friedman, Lisa Gandy, Identifying painting genre
using neural networks. Northwestern University.
G. Y. Chen and B. Kegl. Feature Extraction Using Radon, Wavelet
and Fourier Transform. Systems, Man and Cybernetics, 2007. ISIC.
IEEE International Conference on, pp. 1020-1025. Oct. 2007.
Rafael C. Gonzalez, Richard E. Woods. Digital Image Processing.
2nd edition. Prentice-Hall. 2002.

similar documents