SGN-45007 Computer Vision, 5 cr

Additional information

This course provides an introduction to computer vision including fundamentals of image formation & filtering, feature detection & matching, structure-from-motion & image-based 3D modelling, motion estimation & tracking, and object detection & recognition. The course gives an overview of algorithms, models and methods, which are used in automatic analysis of visual data. The course is only intended for degree students

Person responsible

Esa Rahtu

Lessons

Implementation Period Person responsible Requirements
SGN-45007 2019-01 3 Esa Rahtu

Learning Outcomes

Students will learn - Image formation and image processing - Feature extractions - Object detection and tracking - Use of multiple images - Mapping, visual reconstruction, VR, AR - SLAM - Deep NN

Content

Content Core content Complementary knowledge Specialist knowledge
1. Image formation     
2. Image processing     
3. Feature detection & matching     
4. Feature based alignment & image stitching     
5. Dense motion estimation     
6. Structure from motion     
7. Stereo and 3D reconstruction     
8. Recognition     

Instructions for students on how to achieve the learning outcomes

You must actively participate the lectures and do the exercises. In particular, familiarize yourself with the exercise questions before the exercise session.

Assessment scale:

Numerical evaluation scale (0-5)

Partial passing:

Completion parts must belong to the same implementation

Additional information about prerequisites
Programming skills and basic knowledge of data structures and mathematics (linear algebra, probability) will be necessary.



Correspondence of content

Course Corresponds course  Description 
SGN-45007 Computer Vision, 5 cr SGN-45006 Fundamentals of Robot Vision, 5 cr  

Updated by: Rahtu Esa, 31.03.2020