Back to Projects

Image Segmentation: Breaking it down

Wednesday 2nd November 2016

Image Segmentation is a prominent field in Computer Vision and has a broad range of practical applications: from MRI scans, to aesthetic image quality analysis, to pedestrian detection in video surveillance. This talk will introduce the problem and give an example-driven introduction to some of the many approaches for solving it. We will start by exploring a few relatively simple binary thresholding techniques, before looking at the K-Means clustering algorithm applied to images, and lastly thinking about graph-based methods.

Click here for the talks.cam page, or click below for a 2-up copy of the talk slides.

For a copy of the demo, click below.

For the animated .gifs on page 11, see below.