3D render prototype

$ python3 3drender.py

This script was written for a Linear Algebra evaluation, part of the Data Science and Artificial Intelligence graduation course on FGV - EMAp. There were 10 possible themes to choose for this work, and we (me and github.com/paulocgr) have chosen polygon rendering.

Three-dimensional rendering is a very interesting topic. There are a lot of algorithms on that subject, that goes from basic things like rotating figures to advanced ones like simulation of precise lighting. In this script, we implemented some of the most basic algorithms that turns the 3d rendering possible, most of them that are linear transformations. They are strongly tied to linear algebra, since most of them depends on vectors and matrices (some advanced ones not implemented here depends even on eigenvalues and eigenvectors!).

Some implemented features were scale, translation and rotation for shapes, perspective vision to turn the render scene more simillar to real human vision, backface culling to reduce the amount of rendered figures and so on.

Since the script was written for an university work in Python language, the implementations aren't so economic and fast as they could be, so this isn't meant to be used on any big project. This is to exemplify the importance of linear algebra on the computer 3d rendering.

Libraries

To create the script, only numpy (responsible for efficient implementations of linear algebra concepts) and pygame (responsible for implementing line drawing functions) were used.

Controls

To movement in the world space, the W, A, S, D, up and down keys can be used for translation, N, M, J, K, left and right keys can be used for rotating the camera along three different axis and U and I keys can be used for changing the distance of the near plane used for perspective (a.k.a changing FOV)