Detecting isolated region in a heatmap? [closed]
I am a beginner to image analysis and am trying to write a program that does the following-
What I have is the image of a heat map. Something that looks like 
What I need to do is for the program to detect if there is any region of 'problem'. In this case red would be considered problem. I went through tutorials and know how to detect if the color of region of problem is already known. I need to the program to tell me if there is a problem even if the two colors(red and green) is swamped. In that case, the problem would be green. So basically, if there is a patch of a color anywhere within a larger region of a different color, the program should draw a countour around the inner region. Any suggestion would be appreciated.
PS. I am using opencv 3 with python 2.7 running on raspbian jessie