How does detectCheckerboardPoints work?

Hi!

I am wondering how the function "detectCheckerboardPoints" works. What prerequisites are there for the image I? What algorithm is behind the function? E.g., does it rely on a harris corner detector? Does it explicitly look for quadrilaterals? Why doesn't it work for this image, for example?

Answer 1

[Image Analyst]

Probably because it's such a poor quality image. You might have to do some preprocessing first to get rid of the noise and posterization, and of course better image capture environment (lighting, geometry, etc.).

[Jdeen]

well the image itself has a rather nice quality, the image I copied is just the part of the original image containing the chessboard. I did not want to show the whole scene ;-)

Could you either elaborate on the preprocessing you think needs to be done or on how the function "detectCheckerboardPoints" works? Then I might accept your answer :)

[Image Analyst]

Sorry, but I don't have the toolbox with that function so I don't know about it and can't try it. How did your image get posterized in the first place? It looks like some strange colormap was applied. Don't do that. That's the main problem you need to avoid.

Next, if it's the checkerboard that you want to see, see if you can point your camera at it perpendicularly, and make sure it has enough illumination to get the full 0-255 range, that the illumination is uniform (use a light booth if you can) and make sure it's in focus.

[Jonathan Rudd]

Why answer when you don't know and don't have the toolkit.

[Image Analyst]

Because if it's important enough for him, he might buy the toolbox. Why would I not refer him to an answer just because I personally don't have the toolbox that has a function I know would help him? Why should I withhold such potentially valuable and helpful information? (By the way, I have the toolbox now.)