Introduction

In the previous chapter, we learned about histogram equalization, which was used to enhance an image by bringing out the hidden details in a dark image. In this chapter, we will learn how to fetch objects of interest from an image. We will start with a gentle introduction to contours and then move ahead to see some interesting ways in which you can apply this concept. A contour is the boundary of an object – it is a closed shape along the portion of an image that has the same color or intensity. Speaking in terms of OpenCV, a contour is the boundary of a group of white pixels on a black background. Yes, in OpenCV, contours can be extracted only from binary images.

In practical terms, contours can help you to count the number of objects in an image. You can also use contours to identify your object(s) of interest in a given image, for example, to detect a basketball net in an image (Exercise 4.05, Detecting a Basketball Net in an Image). Furthermore, you will...