CS0029 Computer Vision
(AU 2019)
Announcement
Course Description
- Instructor: Ko-Chih Wang (王科植)
- Email: kcwang@ntnu.edu.tw
- Lecture: Wed. 9:10 ~ 12:10 台師大公館校區-綜合館002
- Office hours: TBD
Topics & Slides (tentative)
Part I : Image Formation & Processing
- History of computer vision
- Applications of computer vision
- OpenCV
- Image
- Morphology
Part II : Motion Detection and Tracking
- Feature extraction and histogram of gradient
- PCA
- Optical flow
- Covariance tracking
- Mean Shift tracking
Part III : 3D vision
- Camera model
- Camera calibration
- Image registration
- Stereo vision
- 3D reconstruction
Part IV : Detection and Recognition
- Unsupervised learning
- Supervised learning
- Neural network
- Training of classifier
Assignment
Syllabus
You will learn the CV fundamental techniques in the lectures and implement the methods from scratch in the assignments. At the end of the semester, students should deliver a machine vision-based automatic system.
By the end of this class, you should have sufficient knowledge to:
- understand and learn new computer vision tools.
- continue on advanced artificial intelligence research.
- build automatic systems, such as autonomous robots, abnormal detection systems, or vision-based access control systems.
Grading
Assignment: 60%
Final project: 30%
Attendance: 5%
Discussion: 5%
Academic Misconduct Policy
With regards to your assignments, students are encouraged to discuss with other classmates. However, students should not copy any code from other people. If multiple students really want to turn in the same code segment, you must talk with me before submitting it. With regards to your final project, if students copy any code from websites, you must indicate the code segment that you copied, including the source web address of the code. You are required to contribute most of the code in your assignments and the final project in principle; do not turn in anything mostly written by other people.
I reserve the right to give you a 0% on an assignment if I believe it is clear that you copied most or all of the answers.