Programming Microsoft® Robotics Studio

Integrating Vision

Robotic vision is one area of robotics that has proved to be challenging. Although you can use a Web camera to simply stream images as a bitmap, this is typically not useful if you need the robot to act based on what it observes. It gets even more complicated if you need to identify and categorize specific objects, for example, if you need the robot to distinguish a cat from a dog. Even the task of simple face recognition has proved to be quite complex.

In this chapter, we will build a security monitor service that determines whether movement has occurred. This is accomplished by comparing the frame images captured from a Web camera. If there is a difference between the images, then the service will send an e-mail notification. ...