Optical Character Recognition

Optical Character Recognition (OCR) is one of the favorite topics of research in computer vision and machine learning. There are a lot of efficient off-the-shelf implementations and algorithms readily available for OCR, but for better understanding of the concepts, we will build our own OCR Android application. Before we get down to writing the code for our application, let's take some time to take a look at the different character recognition techniques and how they work. In this chapter, we will use two standard machine learning techniques: k-nearest neighbors (KNN) and Support Vector Machines (SVM), while building our applications.

The aim of this chapter is to build a real-time digit recognition application. The application will have a live camera output being displayed on the mobile screen and as soon as the camera captures a digit, we will recognize the digit.

OCR using k-nearest neighbors

k-nearest neighbors is the one of the simplest algorithms used for...