Computer Vision

• Identify relevant methods for solving computer vision problems.
• Implement a chosen computer vision algorithm in e.g. Matlab or Python.
• Carry out a systematic performance analysis of a computer vision algorithm.
• Apply one and two view geometry for estimating points, positions, and surfaces.
• Implement and use linear methods for camera estimation.
• Implement and use camera calibration.
• Implement and apply the RANSAC algorithm.
• Find correspondences between 2D image points and estimate 3D points from these.
• Understand and use image features in computer vision.
• Use common computer vision software libraries.

The course covers computer vision methods based on a combination of geometry, mathematics, statistics, and learning-based approaches, applied to images and visual data. Students are introduced to both fundamental concepts and practical methods in computer vision, and gain hands-on experience through exercises involving implementation, experimentation, and evaluation. The course includes both classical methods and selected modern approaches. Students are expected to have prior experience with Python programming, basic image analysis, and introductory deep learning.

02456/02516, Experience programming in Python and NumPy. Image analysis or similar introductory course covering fundamental image representations and operations, including filtering. Familiarity with deep learning concepts and frameworks is also required

Lectures and exercises

Minimum: 6.

Please be aware that this course will only be held if the required minimum number of participants is met. You will be informed 8 days before the start of the course, whether the course will be held.