How to update a background model in motion detection

Hi I am implementing an equation for my abckground model. I have its mathematical form and i believed i implemented it on MTALAB but there seems to be a problem. Either there is some thing missing in my synatx so i was wondering if any one out there can ahve a lookl and correct mr or guide me. Basically what it does is the if equation get satisifes it fills 1on that location (1 in image pixel) else it put 0 in that location.This is my equation

imageGray-meanofIMage>3max(imageVariance,cameranoiseVariance)

if true
   if(abs(imGray(row_loop,col_loop)-state.meanIm(row_loop,col_loop))>(3*max((state.varIm(row_loop,col_loop)),state.cameraNoiseVar)))
   relutantIm=1;
else
   resultantIm=0;
end

Now what i recieve is image that contains all zeros where as it should have some 1 i am sure about that.My values are right since i have a refrence to cross check it i believe something is wrong with my syntax.

Any helps are appreciated

Answer 1

[Image Analyst]

You want resultantIm to be an entire image, right? Not just one number? So do this:

resultantIm = abs(imGray-state.meanIm) > 3*max(state.varIm,state.cameraNoiseVar);

No need for looping over row_loop and col_loop. You can do it all in one line. resultantIm is a logical (binary) image - 0 or 1, or false or true.

To get the masked foreground

foregroundImage = imGray; % Initialize
foregroundImage(~resultantIm) = 0; % Blacken background.

To get the background image:

backgroundImage = imGray; % Initialize
backgroundImage(resultantIm) = 0; % Blacken foreground.

[Lora]

Thanks a lot really