In this chapter, we learned how to analyze simple spatial relationships within images so that we can differentiate between multiple objects, or between a foreground and a background. Our techniques included extraction of three-dimensional information from a two-dimensional input (a video frame or an image). First, we examined depth cameras, and then epipolar geometry and stereo images, so we are now able to calculate disparity maps. Finally, we looked at image segmentation with two of the most popular methods: GrabCut and Watershed.
As we progress through this book, we will continue to extract increasingly complex information from images. Next, we are ready to explore OpenCV's functionality for detection and recognition of faces and other objects.