Now it is time for us to get into a really neat application. We'll start off by using the open source package http://www.aforgenet.com/ to build a face and motion detection application. To do this, you'll need to have a camera installed in your system to see live streaming video. From there, we will use that camera to detect faces as well as motion. In this chapter, we are going to show two separate examples: one for facial detection, the other for motion detection. We'll show you exactly what goes on, and just how fast you can add these capabilities into your application.

In this chapter, we will cover such topics as:

• Facial detection
• Motion detection
• How to use the local video-integrated camera
• Image filtering/algorithms

Let's start out with facial detection. In our example, I'm going to use my friendly little French bulldog to pose for us. Before I do that, please re-read the chapter title. No matter how many times you read it, you'll probably miss the key point here. Notice it says face DETECTION and not face RECOGNITION. This is so very important I wanted to stop and re-stress it. We are not trying to identify Joe, Bob, or Sally. We are trying to verify that, out of everything we see via our camera, we can detect that there is a face there. We are not concerned with whose face it is, just the fact that it is a face! It is so important that we understand this before moving on, otherwise your expectations will be so incorrectly biased that you are going to make yourself confused and upset, and we don't want that!

Facial detection, as I will stress again later, is the first part of facial recognition, a much more complicated beast. If you can't identify, out of all the things that are on the screen, one or more faces, then you'll never be able to recognize whose face it is!