We started this chapter by learning about a template-matching algorithm and an object detection algorithm that, despite its popularity, lacks some of the most essential aspects of a proper object detection algorithm, such as scale and rotation invariance; moreover, it's a pure pixel-based object detection algorithm. Building upon that, we learned how to use global maximum- and minimum-detection algorithms to interpret the template-matching algorithm result. Then, we learned about corner- and edge-detection algorithms, or in other words, algorithms that detect points and areas of significance in images. We learned how to visualize them, and then moved on to learn about contour-detection and shape-analysis algorithms. The final section of this chapter included a complete tutorial on how to detect keypoints in an image, extract descriptors from those keypoints, and use matcher algorithms to detect an object in a scene. We're now familiar with a huge set of algorithms that can be used to analyze images based not only on their pixel colors and intensity values, but also their content and existing keypoints.

The final chapter of this book will take us through computer vision and machine learning algorithms in OpenCV and how they are employed to detect objects using a previously existing set of their images, among many other interesting artificial-intelligence-related topics.