képkockakülönbség

Képkockakülönbség, often translated as "frame difference," is a fundamental technique in computer vision and video processing used to detect changes between consecutive frames in a video sequence. The core principle involves subtracting the pixel values of one frame from the corresponding pixel values of the next frame. Where significant differences exist, it indicates motion or some form of change within the scene.

The process typically begins by acquiring two successive frames from a video stream. These frames are then registered, meaning they are aligned to compensate for any minor camera movement or perspective shifts. Following registration, a pixel-wise subtraction is performed. Pixels that remain relatively unchanged will have a difference close to zero. Pixels that have changed will exhibit a non-zero difference. A threshold is often applied to this difference image to distinguish between genuine changes and noise. Pixels exceeding the threshold are flagged as differing.

Applications of frame differencing are widespread. It is a primary method for motion detection, a crucial step in object tracking, video surveillance for identifying intruders or unusual activity, and in video compression to identify areas that do not require re-encoding. While simple and computationally inexpensive, basic frame differencing can be sensitive to lighting variations and camera noise. More advanced techniques often incorporate background subtraction or optical flow to refine the detection of motion.