Comments on: Line Segment Intersection

By: Rustum Scammell

Rustum Scammell — Tue, 11 Oct 2011 09:17:30 +0000

Jesse, this is great stuff. Combined with the separating axis theorem, it’s the beginning’s of a good collision detection and response system. I’d like to make a small correction to the math above. The factor (x1-x3) in the final solution for Ua and Ub, should be (x3-x1). Cheers!

By: Jesse

Jesse — Tue, 23 Feb 2010 18:06:19 +0000

Tobias, thanks for commenting. I’m glad it was helpful.

It appears you’re correct, that’s definitely a flaw in my code. I’ll write an update when I get a chance.

By: Tobias Wehrum

Tobias Wehrum — Tue, 23 Feb 2010 17:49:30 +0000

First of all, thank you for your code and the explaination! It was very helpful.

This being said, there is one problem with it: When two line segments are coincident, but don’t overlap, your code says they intersect anyway.
This is also described (with a provided solution) at http://stackoverflow.com/questions/2255842/detecting-coincident-part-of-two-line-segments-in-c

By: TyphoidHippo

TyphoidHippo — Sat, 23 May 2009 01:28:56 +0000

This is the only place on the entire internet I could find an understandable, and useful example of how to do this. I’m actually using C++ and OpenGL at the moment, but the clear and concise logic in your XNA/C# code translated with ease. Thanks so much for sharing this, seriously – I was about to pull my friggin hair out.

By: Ryan

Ryan — Wed, 17 Dec 2008 00:41:07 +0000

I’m trying a very simple Breakout clone using XNA at the moment. It will be my first ever full game.

XNA is fantastic but eventually I want to start working with C++ and DirectX.

By: Jesse

Jesse — Tue, 16 Dec 2008 23:44:42 +0000

Hit me with the URL when you get your blog up. I’m up to about six million game development blogs that I follow, but always looking for more. I don’t know what I’d do without RSS.

What sort of stuff are you working on/with?

By: Ryan

Ryan — Tue, 16 Dec 2008 23:39:24 +0000

Thanks for the quick reply, Jesse. I’m a Computer Science student majoring in Games, and consider myself a hobbyist as well.

I might be starting my own blog soon, and would love if we could be link buddies. Love your blog.

By: Jesse

Jesse — Tue, 16 Dec 2008 14:50:09 +0000

That’s exactly right. When dealing with floating point numbers, you have to recognize that they have limited precision. The denominator is created by subtracting and multiplying eight floating point numbers, so instead of testing to see if it’s exactly zero, we just test if it’s pretty close.

By: Ryan

Ryan — Tue, 16 Dec 2008 06:30:42 +0000

I have a question. Why do you use

if (Math.Abs(denominator) <= 0.00001f)

instead of

if (Math.Abs(denominator) == 0f)? A tolerance of somekind?

Great post, by the way.

By: Simon (Darkside) Jackson

Simon (Darkside) Jackson — Wed, 12 Nov 2008 10:42:29 +0000

Great stuff, it always amazes me how you transform the mathematical notation into actual code in C#.
Always a lot easier to read for math noobs like myself (I’ve forgotten most of my math from school, so many years ago )