Hello guys,

I have a question, could anyone please guide me?

In one research, I want to detect the license plate. I find a lot of research about that. But, in those projects, the video camera is constant and the car moves. In my research, the video camera is moving and the cars are constant. My system should detect the license plate lower than .1 seconds for each frame (I only want to detect the license plate and don't want to recognize them at this time). In other words, because the video camera moves, there are a lot of frames which don't have license plate inside them and there are some frames which consist license frame. My system should recognize which frames have the license plate and which don't. This process for each frame should take about .1 second. Could anyone guide me which method is suitable?

Thanks