Episode 5: Computer vision with opencv

In this episode we introduce opencv, a popular open source computer vision library created by Dr. Gary Bradski and developed in large part by a team of Russian computer vision and optimization experts. We interviewed Gary at his office at Willow Garage, where they are building an open research platform for personal robotics, including the PR2 above.

Handy links below.

General

• OpenCV, the official page.
• Learning OpenCV, a very thorough introduction from O’Reilly publishing, written by Dr. Gary Bradski and Dr. Adrian Kaehler.
• Sarov, the secret Soviet city (still apparently closed to foreigners).
• How does Sudoku Grab work? An explanation of the iPhone Sudoku solver Nat uses to cheat in multiplayer games online.
• CubeCheater, the iPhone Rubik’s Cube solver which is now offline due to legal threats (!!). But you can still see a video.
• TinEye, the reverse image search engine. Find out where an image comes from.
• Google Goggles.
• The Poseidon Drowning Detection System.

DARPA Grand Challenge

• Wikipedia’s overview of the DARPA Grand Challenge
• Stanley, the car from Stanford which won in 2005.
• A nice video recap of the 2005 DARPA Grand Challenge
• Demo of Stanley’s vision system

OpenCV Hacks

• ASL Finger Recognition using C# OpenCV wrapper
• Tracking a tennis ball
• Control your mouse in Linux with a red glove
• Face features detection system
• Toss and catch a virtual ball
• Use your head as a mouse with MouseTrap
• Sample code for OpenCV face detection
• Sample code for grabbing and displaying frames from a webcam
• Emgu, a .NET wrapper for OpenCV.

Transcript

00:00:00 00:00:13 Gary: It makes the little line matching canal find all the triangulations to fill in this scene in a dense way. If you have dense, nice 3D models then there’s a lot you can do in perception. But it also works – it’s getting – you can’t see this but this is rapidly flashing on and off and so it gets a clear image and then it gets the dense image and then the other stereo comes in between all that and gets a color image. And then there’s this camera, too, is a high resolution five megapixel, but that only goes … 00:00:44 00:00:52 Nat: That was Gary