Course Title : Spacio-Temporal Media Analysis

Code 10C714
Course Year Master Course
Term 1st term
Class day & Period Tue 3rd
Location Yoshida campus(N1)・Katsura campus(A1-131)・Ujicampus(S-143)
Credits 2
Restriction No Restriction
Lecture Form(s) Lecture
Language Japanese or English
Instructor Yuichi Nakamura, Kazuaki Kondo

Course Description

Representation, feature extraction, recognition of media with two or higher dimensions, especially images and videos, are explained with comparing to human vision and biological systems.


Evaluation is based on participation and reports.

Course Goals

To learn the basic of representation, feature extraction, and pattern recognition of signals with two or higher dimension, and their applications.

Course Topics

Theme Class number of times Description
Spatio-Temporal Media 1 What is spatio-temporal media. Some examples.
Light and Colors 1-2 Intensity, colors, and spectrum in image media.
Features and Segmentation 2 Features such as edge, region, etc. for analysing image media.
Filtering and Wavelet Transform 1-2 Introduction to filtering and Wavelet Transform.
Discrete Wavelet Transform and Applications 1-2 Dicrete Wavelet Transform and applications such as image enhancement, image compression, etc.
Geometry of Image Capturing 1-2 The mechanism and geometry of image capturing: projection of a 3D world into 2D images.
3D Measurements and Reconstruction 2 3D measurements and 3D world reconstrunction from a set of 2D images.
Measurement of Motions 1-2 Motion detection and measurement, and oject tracking.
Pattern Recognition 0-2 The basic idea of pattern recognition and usuful tools such as Support Vector Machine.


No specific textbooks. Handouts will be given when necessary.


Computer Vision: A Modern Approach, Forsyth and Ponce, Prentice Hall


Fundamental knowledge of digital signal processing

Independent Study Outside of Class

Web Sites

Please see PandA (

Additional Information