Why do image errors occur by using a webcam and a Simulink-model on a raspberry pi?

Hello, I am using a webcam (LifeCam Studio, Microsoft) and a Simulink-model running on a Raspberry Pi model B to count moving objects (similar to the example "Analyzing Traffic Using a Webcam, a Raspberry Pi and ThingSpeak"). By increasing the resolution, I can identify more image errors (parts of the image are mixed up and not in the right position) and the counting doesn't work properly. Does anyone please know why this happens and how to solve this problem?