Page 1 of 1

Making an App Compatible With All Cameras and Environments

PostPosted: Mon Jan 14, 2019 4:23 pm
by njsokalski
Although I have written many .NET apps, this is my first app using AForge or computer vision or image analysis, so I am very unfamiliar with certain terms used in these areas. I am attempting to write an app that is used to recognize a card containing a 5x5 grid of colored squares, each of which is 1 of 4 colors. I have managed to get this to work on one machine using filters and blobs, but I am still having trouble getting it to work on my other test machine. I am assuming this is because all machines have different cameras and because differences in the environment (such as lighting). I completely clueless as to what to do next to try and figure out how to make the app work for all (or most) environments. I am not expecting people to write my code for me, I am simply asking if anybody knows of any good places or sites with good information on how to solve this kind of problem. Any recommendations? Thanks.

Re: Making an App Compatible With All Cameras and Environmen

PostPosted: Tue Jan 15, 2019 8:36 am
by andrew.kirillov
If your application is based on color recognition, then I think it will very difficult to make it work in every possible environment. As you already mentioned, depending on light conditions color values will vary a lot. You may try using different color spaces, other than RGB. For example, you may try HSV/HSL. You may research further and find other color spaces, which are more invariant to light conditions. But still, with things like shadows, reflections, etc. - will not be easy.

I don't know what your application is. But it might be easier to look for some features, instead of colors. What about something like glyphs?