The video analysis module in OpenCV is a collection of extremely powerful algorithms, functions, and classes that we have learned about in this chapter. Starting from the whole idea of video processing and simple calculations based on the content of consecutive video frames, we moved on to learn about the Mean Shift algorithm and how it is used to track objects with known colors and specifications using a back-projection image. We also learned about the more sophisticated version of the Mean Shift algorithm, which is called the Continuously Adaptive Mean Shift, or simply CAM Shift. We learned that this algorithm is also capable of handling objects of different sizes and determining their orientation. Moving on with the tracking algorithms, we learned about the powerful Kalman filter and how it is used for de-noising and correcting the tracking results. We used the Kalman filter to track mouse movements and to correct the tracking results of the Mean Shift and CAM Shift algorithms. Finally, we learned about OpenCV classes that implement background-segmentation algorithms. We wrote a simple program to use background-segmentation algorithms and output the calculated background and foreground images. By now, we are familiar with some of the most popular and widely used computer vision algorithms that allow real-time detection and tracking of objects.

In the next chapter, we'll be learning about many feature extraction algorithms, functions, and classes, and how to use features to detect objects or extract useful information from images based on their key points and descriptors.