Advanced Image Processing
Lecture: Thursday, 14:15-16:00, Battelle (Bat 316-318) ( beginning February 19 )
Labs: Wednesday, 14:15-16:00, Battelle (Bat 314-315) (beginning February 18)
CUI - University of Geneva, Computer Vision Multimedia Laboratory
If you are interested in this course and wish to be added to mailing list, please, send an
- Sviatoslav Voloshynovskiy
(Office 215, Office hours: Tuesday 11.00-12.00 or according to appointment by email)
- Stephane Marchand-Maillet
(Office 212, Office hours: Tuesday 11.00-12.00 or according to appointment by email)
- Teaching Assistant:
- Taras Holotyak (Office 226)
- Battelle Building A,
- 7, route de Drize,
- 1227 Carouge
- CUI, University of Geneva
- Recall of Linear Algebra. Multidimensional Signal Processing ( ~1.5 lectures)
Vector and matrix image presentations, discrete and continuous Fourier transforms.
- Introduction. Human Visual System ( ~0.5 lecture)
Modulation transfer function, visual masking, noise visibility, color vision;
- Image Representation : pyramids and wavelets ( ~1 lecture)
- Random signals ( ~1 lecture)
- Image Modeling ( ~3 lectures)
Stochastic presentation of images;
Stationary continuous- and discrete-space models, including AR, MRF, stationary Generalized Gaussian;
Nonstationary models: non-stationary Gaussian, HMM;
Transform-based models (DFT, DCT, wavelet);
Edge and texture models;
Doubly stochastic processes;
Relationships between models.
- Image Sensor Models ( ~0.5 lecture)
Optical, radar and medical coherent/noncoherent imaging applications:
(aperture difraction constrains,
defocusing, motion blur, atmospheric turbulence, sparse imaging apertures);
CCD imaging applications;
- Noise Models ( ~0.5 lecture)
Additive noise: Poisson, Gaussian and Laplacian models;
Multiplecative noise: speckle model.
- Image Denoising ( ~1 lecture)
Models selection (MDL principle);
Transform-based denoising: adaptive Wiener filtering, soft-shrinkage and hard-thresholding.
- Image Restoration ( ~1 lecture)
Statistical ill-posed problems;
Deterministic regularization: Tikhonov, edge-preserving and adaptive regularizations;
- Image Compression ( ~1 lecture)
Basics of source coding theory (lossless and lossy);
Vector quantization, codebook design;
Transform and subband coding;
Relationship between compression and denoising.
- Video Modeling and Compression ( ~1 lecture)
3-D and 2-D Motion models;
Block matching (simple, hierarchical, and overlapped), optical flow;
Motion-compensated prediction models;
Transform and motion-based compression techniques;
- Digital Data Hiding ( ~2 lectures)
Steganography (secure communications);
Digital watermarking: fundamentals, channel coding, masking, robustness against geometrical transforms and applications
(robust watermarking, tamper proofing and self-recovering, document authentication, access control, indexing).
(Prof. Thierry Pun), good knowledge of Matlab.
- Supplementary Materials:
- A.K.Jain, Fundamentals of Digital Image Processing, Prentice-Hall, 1989.
- A.M.Tekalp, Digital Video Processing, Prentice-Hall, 1995.
- A.Bovik, Handbook of Image Video Processing, Academic Press, 2000.
- H.Stark and J.W.Woods, Probability, Random Processes, and Estimation Theory for Engineers, Prentice-Hall, 1994.
- A.M.Yaglom, Correlation Theory of Stationary and Related Random Functions I: Basic Results, Springer-Verlag, 1987.
- L.Breiman, Probability, SIAM, 1992.
- H.V.Poor, An Introduction to Signal Detection and Estimation, 2nd Ed., Springer-Verlag, 1994.
- A.Gersho and R.M.Gray, Vector Quantization and Signal Compression, Kluwer, 1992.
- M.Vetterli and J.Kovacevic, Wavelets and Subband Coding, Prentice-Hall, 1995.
- Useful Links: