So, now that you have an idea of how a classification system operates on images, it is time to focus on face detection. In this section, we will give you an idea on the major steps that go into the popular Viola-Jones face detection framework. As we take you through the steps, we will simultaneously also draw parallels with the steps that we have just covered under Image Classification Frameworks. This would give you a good idea as to how these classification systems are implemented in practice (by Viola Jones, in particular).
For ease of explanation, I have divided the framework for Viola and Jones into four major parts, as described as the following:
Haar features
Integral image: an efficient way to compute Haar features
AdaBoost learning: an ensemble of classifiers
Cascaded classifiers: the secret to making the detection framework fast
We will be going through each of these sections one by one, in an attempt to build an understanding of face detection. Before we proceed...
