When we display a 3D object on a 2D image plane, often times the 2D image loses some properties of the 3D object. For example, if it is a 3D wireframe, the lines hidden right behind the lines in the front will not be seen. When the 3D object is very complex such as the veins in the brain, we can't just rotate the object and find a view that looks okay. In this tutorial, we introduce the properties and complexities for having a nonambiguous and nice perspective projection based on [1].

The covered topics here are,

1. Definition of Perspective Projection

2. What is Niceness in projection?

3. Three properties needed to be a Nice Perspective Projection : Regularity, Minimum Crossing, and Monotonicity

The detailed algorithms to take the Nice Perspective Projection are not discussed here.