Creating a human interface in Computer Vision is not an easy task. One of the exciting fields is Head-coupled perspective. This technique is used for rendering the scene on the screen, which responds naturally to changes in the head position of a viewer relative to the display. Simply put, the technology creates a 3D display without using any additional devices except the camera.
In the previous chapter, we saw how to track a head with the headtrackr library. It was done using the Camshift algorithm. In this section, we will explain the background of the function for Head-coupled perspective and how to use it in your projects to create an amazing human interface. To present a scene, the headtrackr library uses one of the most popular JavaScript libraries for 3D modeling—three.js (http://threejs.org). We will begin with an explanation of the core function and then see an example of its usage.
