2020 Computer Vision

Font size  SML

Register update notification mail Add to favorite lecture list
Academic unit or major
Graduate major in Systems and Control Engineering
Instructor(s)
Okutomi Masatoshi 
Class Format
Lecture    (ZOOM)
Media-enhanced courses
Day/Period(Room No.)
Fri7-8(Zoom)  
Group
-
Course number
SCE.I454
Credits
1
Academic year
2020
Offered quarter
4Q
Syllabus updated
2020/9/28
Lecture notes updated
-
Language used
English
Access Index

Course description and aims

The main topic of this course is how to reconstruct 3-D information in space from 2-D images. This course will introduce several cues for reconstruction. Then geometric modeling between space and images, and methods for reconstruction based on the modeling will be explained. In addition, photometric modeling of images and scene reconstruction methods will be introduced.

Student learning outcomes

1. Understanding the outline of 3-D reconstruction from 2-D images.
2. Understanding the mathematical model which represents the geometric relation between space and images.
3. Understanding the methods for "stereo vision" and "structure from motion".
4. Understanding the photometric model of images and the methods for scene reconstruction, such as object shape, surface reflectance, and illumination.

Keywords

3-D reconstruction, geometric model, camera calibration, stereo vision, motion estimation, structure from motion, photometric model, reflectance characteristics

Competencies that will be developed

Specialist skills Intercultural skills Communication skills Critical thinking skills Practical and/or problem-solving skills

Class flow

The outline of 3-D reconstruction from images, its fundamental theories and methods will be explained.
The contents might be changed according to students' knowledge and interests.

Course schedule/Required learning

  Course schedule Required learning
Class 1 Three Dimensional Reconstruction from Images Understanding the essential problem of 3-D Reconstruction from images and the relation with human vision.
Class 2 Geometric Imaging Model and Camera Calibration Understanding geometric modeling to represent space-image relation and camera calibration.
Class 3 Stereo Vision Understanding the principle and equations of stereo vision.
Class 4 Stereo Matching, Multi-View Stereo and Camera Pose Estimation Understanding Stereo Correspondence and Multi-View Stereo for Better Reconstruction. Understanding Camera Pose Estimation.
Class 5 Structure from Motion and SLAM Understanding the problem to estimate both camera motion and 3-D positions.
Class 6 Photometric Imaging Model and Scene Reconstruction Understanding photometric imaging model and methods for reconstructing scene information including illumination, surface reflectance, and shape.
Class 7 Demonstration and Supplement Demonstration of Some Relevant Methods and Supplemental Explanation

Out-of-Class Study Time (Preparation and Review)

To enhance effective learning, students are encouraged to spend approximately 100 minutes preparing for class and another 100 minutes reviewing class content afterwards (including assignments) for each class.
They should do so by referring to textbooks and other course material.

Textbook(s)

We will use the following reference book and handouts.

Reference books, course materials, etc.

Digital Image Processing: Computer Graphics Arts Society (CG-ARTS)

Assessment criteria and methods

The level of understanding about the contents presented in the course and the ability to apply them to problems will be asessed by submitted reports.

Related courses

  • SCE.I301 : Image Sensing
  • SCE.I501 : Image Recognition

Prerequisites (i.e., required knowledge, skills, courses, etc.)

no prior conditions

Page Top