In this chapter, we looked at different image segmentation algorithms, namely, contour detection, superpixels, watershed, and normalized graph cut. These algorithms are fairly easy to implement and run almost real time. Image segmentation has tremendous use in real-world applications like background subtraction, image understanding, and scene labeling. Recent advances in machine learning, especially deep learning, have enabled more sophisticated ways of image segmentation that involve almost no manual tuning of parameters.
In the coming chapters, we will look at some of the machine learning techniques and how they are relevant to computer vision.