161C - Digital Signal Processing II



This course provides an introduction to image processing and computer vision. We start from the basics of image formation (cameras, light, color), then cover the fundamental concepts in 2D signal processing (2D convolutions, Fourier transforms, etc) and low-level vision (edges, texture), and finally move on to higher level problems such as motion analysis, image segmentation, image classification and retrieval.

Lectures: TuTh, 3:30p-4:50p, WLH2208
Instructor: Nuno Vasconcelos
n u n o @ e c e . u c s d . e d u, EBU1-5602
office hours: Friday, 9:30a-10:30a
Teaching Assistant: Hamed Masnadi-Shirazi, h m a s n a d i @ u c s d . e d u
Office hours: Wednesday, 3:00p-4:30p, EBU1-5706
Text: Computer Vision: a modern approach
D. Forsyth and J. Ponce, Prentice Hall, 2003
Secondary text: Two Dimensional Signal and Image Processing
Jae Lim, Prentice Hall, 1990
Syllabus: [ps, pdf]
Homework: Problem set 1 [ps, pdf, data]
Problem set 2 [ps, pdf, data]
Problem set 3 [ps, pdf, data]
Problem set 4 [ps, pdf, data]
Problem set 5 [ps, pdf, data]
Problem set 6 [ps, pdf]
Readings: Lecture 1: introduction [slides, videos]
Lecture 2: cameras (sections 1-1.2.2, 1.3, F&P) [slides, videos]
Lecture 3: radiometry (chapter 4, F&P) [slides]
Lecture 4: radiometry, light sources (sections 5.1-5.2.2, 5.3.1, F&P) [slides]
Lecture 5: color (sections 6.1-6.3.3, F&P)[slides]
Lecture 6: 2D DSP (chapter 1, Lim; chapter 7, F&P)[slides]
Lecture 7: 2D Fourier transforms (chapter 1, Lim; chapter 7, F&P)[slides]
Lecture 8: filtering, smoothing and noise (chapter 8, F&P) [slides]
Lecture 9: edges (chapter 8, F&P) [slides]
Lecture 10: edges, interpolation, templates (chapter 8, F&P) [slides]
Lecture 11: mid-term review [problems]
Lecture 12: mid-term
Lecture 13: 2D DFT (chapter 3, Lim; chapter 7, F&P) [slides]
Lecture 14: DFT (chapter 3, Lim) [slides]
Lecture 15: DCT (chapter 3, Lim) [slides]
Lecture 16: scale, pyramids, and texture (chapter 9, F&P) [slides]
Lecture 17:  least squares [slides] (section 15.2, F&P; section 3.3 Strang)
Lecture 18: motion, least squares [slides] (paper by Lucas and Kanade)
Lecture 19: JPEG [slides]
Lecture 20: MPEG [slides]