What is the parameter that differentiate between static background and moving background when both frame differencing and Mixture of gaussian techniques are used ?

 Trying to model an adaptive background subtraction such that depending on the background model either the frame differencing(for static background) technique is used or Mixture of gaussian model(for dynamic background model) is used.So there must be some parameter that distinguishes between the static and dynamic background.It would be great if someone could help me with this.Thanks.

[Sean de Wolski]

Motion?

[Sean de Wolski]

I think you need to be more precise - perhaps with example images. Otherwise the difference between dynamic and static is simply the presence of motion.

[keethi2]

yeah i agree,presence of motion can be obtained by using the difference.But, " could you suggest me some parameter that would take care of the background clutter and traces as well as that parameter can be used to find if there has been a motion in the image." So,if there has been a motion i could use the Gaussian mixture model else i could use the simplest frame difference technique.

[Sean de Wolski]

Not without seeing an image.

[keethi2]

this basically an algorithm that im tryin to develop,so there are no images at present.But,basically i hope u get my above question ??

[Sean de Wolski]

I understand some of it, but I can't provide you with anything useful without seeing an example. I would hit up the Compendex libraries.

[keethi2]

Im trying to use 2 algorithms adaptively.if i have a static background,i would use frame differenceing technique or if i have a changing background(such as clutter,tree branches waving,illumination change),then i would use mixture of Gaussian.

So,now i need a parameter that would help me determine which of these 2 methods i need to use,with that parameter also taking care clutter,tree branches waving,illumination change. --- hope im clear with the question...thanks