Computer Vision integration (OpenCV,EmguCV)

Hi guys,

Atm I am evaluating possibilites to get OpenCV working within Unity on IOS and Android devices.I want to be able to have some object detection. I don’t know if there will be some performance issues on the devices since most of the samples I saw were using none-mobile devices. I was looking mostly at the Emgu CV and opencvsharp wrapper implementation but as it seems, one problem is the missing System.Drawing namespace.Apparently there is one guy which got this working somehow but I am not sure if this is a proper approach.

As far as I know the OpenCV library is the only library which can fullfill the requirements I need (detection of small colored sphere in relation of a game field/position)

I was also thinking of using the metaio unity plugin and for the colored sphere detection I use some other algorithm with process the camera picture (blob detection). I am also wondering if there are unity plugins which provide some algorithm for static images like blob detection or hough circle transform or any other computer vision algorithm which can be used on images.

I am wondering if you guys have some helpful suggestions or idea you would like to share with me.

Any help is highly appreciated.

Instead of using .NET OpenCV Library ie. Emgu CV it is better to write down Unity 3d Plugins for iOS or Android Separately because it is free & easily integrated .In addition to that you easily found EXAMPLE how to integrate OpenCV library in Android or iOS on github?

I made Unity plugin for iOS to access OpenCV library functions for face detection.

I solved this a while ago by taking the System.Drawing DLL file and placing it in the Assets folder. See this. However, I think I was using OpenCV, since it was for a Windows/standalone game. In your case, you should use EmguCV, and be sure to take Mono’s System.Drawing DLL file so it works on iOS. This file might be included in EmguCV’s source code. Good luck!

Here is something my colleague made :