Detect Quadrilateral points in high contrast image

Question

I need to detect points of quadrilateral in a pretty high contrast image. I understand how I can detect large changes in contrast between 2 pixels, but I'm wondering what would be the best way to detect entire boundaries and corners of a quad in an image.

So I'm basically looking for a good article/algorithm which explains/does this. Note I've seen articles which detect edges but don't actually turn these into vector-based lines. It's the corner points I'm really after! :)

Answer 1

Have a look at AForge- it's got great computer vision capabilities that you can build on, and it's open source to boot, so even if it doesn't do what you want out of the box, you can get some ideas.

Answer 2

The Hough Transform is a very useful algorithm for your task. Here are a few links: 1) wikipedia, 2) more detailed with examples -- but on solid shapes, 3) an example using points.

Answer 3

Use Corner detection techniques, like Harris's or SUSAN. OpenCV could help you.