We have previously studied how to describe objects by means of key points and features, and how to find the correspondence points in two different images of the same physical object. However, our previous approaches were rather limited when it came to recognizing objects in real-world settings and assigning them to conceptual categories. For example, in Chapter 2, Hand Gesture Recognition Using a Kinect Depth Sensor, the required object in the image was a hand, and it had to be nicely placed in the center of the screen. Wouldn't it be nice if we could remove these restrictions?
The goal of this chapter is to train a multiclass classifier to recognize traffic signs. In this chapter, we will cover the following concepts:
- Planning the app
- Briefing on supervised learning concepts
- Understanding the German Traffic Sign Recognition Benchmark (GTSRB) dataset
- Learning about dataset feature extraction
- Learning about support vector machines (SVMs)
- Putting it all together
- Improving results with neural networks
In this chapter, you will learn how to apply machine learning models to real-world problems. You will learn how to use already available datasets for training models. You will also learn how to use SVMs for multiclass classification and how to train, test, and improve machine learning algorithms provided with OpenCV to achieve real-world tasks.
We will train an SVM to recognize all sorts of traffic signs. Although SVMs are binary classifiers (that is, they can be used to learn, at most, two categories—positives and negatives, animals and non-animals, and so on), they can be extended to be used in multiclass classification. In order to achieve good classification performance, we will explore a number of color spaces, as well as the Histogram of Oriented Gradients (HOG) feature. The end result will be a classifier that can distinguish more than 40 different signs from the dataset, with very high accuracy.
Learning the basics of machine learning will be very useful for the future when you would like to make your vision-related applications even smarter. This chapter will teach you the basics of machine learning, on which the following chapters will build.