Edge detection and Corner detection are two of the most basic feature detection algorithms, and are very useful ones too. Having information about the edges in an image can be of great help in applications, where you want to find boundaries of different objects in an image, or you need to find corners in an image when you want to analyze how an object rotates or moves in a given sequence of images (or videos). In this section, we will take a look at the techniques and implementations of various Edge and Corner detection algorithms, such as Difference of Gaussian, Canny Edge detector, Sobel Operator, and Harris Corners.

So far, we looked at how to detect edges and corners in an image. Sometimes, for image analysis apart from edges and corners, you want to detect shapes, such as lines, circles, ellipses, or any other shape for that matter. Say for example, you want to detect coins in an image, or you want to detect a box or a grid in an image. A technique that comes handy in such scenarios is Hough transformations. It is a widely used technique that detects shapes in an image using their mathematical equations in their parameterized forms.

The generalized Hough transformation is capable of detecting any shape for which we can provide an equation in the parameterized form. As the shapes start getting complex (with an increase in the number of dimensions), such as spheres or ellipsoids, it gets computationally expensive; hence, we generally look at standard Hough transformations for simple 2D shapes, such as lines and circles.

In this section, we will take a look at Hough transformations...

We are often required to break down the image into smaller segments to have a more focused view of the object of interest. Say for instance, you have an image with balls from different sports, such as a golf ball, cricket ball, tennis ball, and football. However, you are only interested in analyzing the football. One way of doing this could be by using Hough circles that we looked at in the last section. Another way of doing this is using contour detection to segment the image into smaller parts, with each segment representing a particular ball.

The next step is to choose the segment having the largest area, that is, your football (it is safe to assume that the football would be the largest of all!).

Contours are nothing but connected curves in an image or boundaries of connected components in an image. Contours are often computed using edges in an image, but a subtle difference between edges and contours is that contours are closed, whereas edges can be anything. The concept of...

Let's try to apply our learning from this chapter and create a simple application for detecting a Sudoku grid in an image. You see Sudoku puzzles every day in newspapers or magazines; sometimes they provide a solution and sometimes they don't. Why not build an application for your mobile phone that can click a picture of the Sudoku, analyze the numbers, run some intelligent algorithms to solve the puzzle, and within seconds, you have the solution on your mobile screen.

After reading this chapter, we can easily work on the first part of the application; that is, localizing (detecting) the Sudoku grid from an image. As you read through this book, you will come across algorithms and techniques that will help you build the other parts of this application.

Let's break down our problem statement into subproblems and work each one out to have a fully functioning Sudoku localizing application: