Apply for Scholarship
Computational Vision (CMPT 412)
Computer vision is the process of automatically extracting information from images and videos. The course covers various aspects of Computer Vision, for example, imaging geometry (camera calibration, stereo, and panoramic image stitching), video analysis (motion detection and tracking), image segmentation, object recognition, and more. The course teaches both traditional techniques and more recent learning-based approaches such as deep neural networks, while we will focus increasingly more on the latter. The course will be based on lectures and assignments (Python and Matlab). Computational approaches to image understanding will be discussed in relation to theories about the operation of the human visual system and with respect to practical applications in robotics. Topics will include edge detection, shape from shading, stereopsis, optical flow, Fourier methods, gradient space, three-dimensional object representation and constraint satisfaction.