EE 264: Image Processing and Reconstruction
Fundamental concepts in digital image processing and reconstruction. Continuous and discrete images, image acquisition, sampling. Linear transformations of images, convolution and superposition. Image enhancement and restoration, spatial and spectral filtering. Temporal image processing: change detection, image registration, motion estimation. Image reconstruction from incomplete data. Applications.
||Engineering 2 , Room 243A
||T/Th 10:00 to 11:45, Baskin 156
||T/Th 1-2 PM or by appointment
||Digital Image Processing by Gonzalez and Woods, Third Edition (Errata)
|Optional Reference Texts:
- Digital Image Processing Using MATLAB by Rafael C. Gonzalez, Richard E. Woods, Steven L. Eddins
- Digital Image Processing by William K. Pratt
- Fundamentals of Digital Image Processing by Anil K. Jain
- Two-Dimensional Imaging by Ronald N. Bracewell
||Homeworks (20%), Midterm (30%), Final Project (50%)
|Notes on MATLAB:
||Homework exercises will require the use of the software package MATLAB. Here are some resources to get you started:
|First day of class
||Thursday, Sept. 22
|Last day of class
||Thursday, December 1
||Tuesday, November 8
|Final Project presentations
||Tuesday, December 6
|Final Project reports, presentation slides, and code due
||Tuesday, December 6
- Homework 1: due Tues Oct. 4 (Solutions)(imgfilter, spatfilt, InterpBilinear)
- Text problems 2.7, 2.19, 3.11, 3.29, 3.30
- MATLAB projects 02-04, 03-04, 03-05
- Homework 2: due Thurs. Oct 13 (Solutions) (Matlab 4.03-04)
- Text problems 3.23 , 3.27, 4.20a,c,e, 4.21, 4.22, 4.27, 4.28, 4.39
- MATLAB projects 4-03, 4-04
- Homework 3: due Thurs. Oct. 27 (Solutions)(Matlab 5.03, Matlab5.04, Example of SD reconstruction)
- Text problems 4.29, 4.30, 5.17, 5.21, 5.22, 5.26
- MATLAB projects 5-03, 5-04 (+part(e): Restore the image using Steepest Descent method)
- Project Summary: due Thurs. Nov. 3 according to the guidelines here.
- Projects will be on a topic from the list (link below), with my approval
- Project teams of at most 2 students
- SELECT YOUR PROJECT ASAP, but no later than the midterm date. Only one team will be able to work on each project topic.
- Project deliverables:
- Written report approximately 15 pages long. PDF format. (Example)
- 15-20 minute oral presentation. PPT/PDF format. (Example)
- Implementation of project ideas in computer code: MATLAB format in Zip Archive. (Example)
- Final Projects Page (UPDATED WITH REQUIRED PROJECT SUMMARY DESCRIPTION)
Links and Matlab Demos:
Tentative Syllabus and Reading:
- Week 1 (Ch. 1)Introduction, overview of history and applications of Image Processing. Review of the requisite mathematical concepts, including concepts in matrix, probability, and statistical analysis.
- Weeks 2 (Ch. 2)Fundamentals of Image Processing. Visual Perception, light and the EM spectrum. Image sensing and acquisition. Sampling and quantization.
- Weeks 3/4 (Ch. 3)Spatial domain image enhancement. Gray-level transforms, histogram procesing, averaging and subtraction of images. Retinex. Spatial Filtering: Convolution and superposition. Smoothing/sharpening filters.
- Week 5 (Ch. 4)Spectral domain enhancement. Linear transformations of images. Fourier, DCT, etc. Smoothing/sharpening filters. Homomorphic filtering.
- Week 6/7 (Ch. 5)Image Restoration. Modeling the degradation process. Noise and interfering patterns. Linear, space-invariant degradations. Blind restoration/estimation of the degradation process. Inverse filtering. Wiener filter, Least-squares filter. Geometric transfromations.
- Week 7 (Ch. 7)Wavelets and Multiresolution processing. Image pyramids, Continuous and discrete wavelet transforms. Wavelets in 2 dimensions. Applications to compression, enhancement, restoration.
- Week 8 (Ch. 9/10)Basic Image Analysis. Detection of discontinuities and change in images, segmentation and thresholding. Basic morphological processing. Description and respresentation of shape.
- Week 9 (Notes)Advanced Topics. Multiple-image processing. Image registration. Motion estimation. Resolution enhancement. Inverse Problems in imaging.
- Week 10 (Notes)Review of material, presentation of final projects.
- EE 262: Statistical Signal Processing
- Covers fundamental approaches to designing optimal estimators and detectors of deterministic and random parameters and processes in noise; includes analysis of their performance. Binary hypothesis testing: the Neyman-Pearson Theorem. Receiver operating characteristics. Deterministic versus random signals. Detection with unknown parameters. Optimal estimation of the unknown parameters: least square, maximum likelihood, Bayesian estimation. Reviews the fundamental mathematical and statistical techniques employed. Many applications of the techniques are presented throughout the course.
- EE 264: Image Processing and Reconstruction
- Fundamental concepts in digital image processing and reconstruction. Continuous and discrete images, image acquisition, sampling. Linear transformations of images, convolution and superposition. Image enhancement and restoration, spatial and spectral filtering. Temporal image processing: change detection, image registration, motion estimation. Image reconstruction from incomplete data. Applications.
- EE 153: Digital Signal Processing
- Introduction to the principles of signal processing, including discrete-time signals and systems, the z-transform, sampling of continuous-time signals, transform analysis of linear time-invariant systems, structures for discrete-time systems, the discrete Fourier transform, computation of the discrete Fourier transform, filter design techniques.
- EE 103: Signals and Systems
- Characterization and analysis of continuous-time signals and linear systems. Time domain analysis using convolution. Frequency domain analysis using the Fourier series and the Fourier transform. The Laplace transform, transfer functions and block diagrams. Continuous-time filters.
Updated July 2012
Peyman Milanfar is a Professor of Electrical Engineering (Currently on leave at Google.) He was Associate Dean for research and graduate studies from 2010 to 2012. He received the B.S. in EE/Mathematics from Berkeley, and his Ph.D. in EECS from MIT. Prior to UCSC, he was at SRI, and a Consulting Professor of CS at Stanford. In 2005 he founded MotionDSP , which has brought state-of-art video enhancement to market. His interests are in statistical signal, image and video processing, computational photography and vision. He is a member of the IEEE Signal Processing Society’s IVMSP Technical Committee, and its Awards Board. He is a Fellow of the IEEE.