A raytracer built from scratch in c++ using the OpenCV and Eigen libraries. Featuring shadow, refraction and reflection rays (using the hall shading model) as well as point lights, spot lights, polygonal meshes and spheres, limited 2D texture map support, adaptive supersampling, and the bounding volume heirarchy acceleration technique.

Images from each step of building the raytracer: https://www.cs.drexel.edu/~ae588/index.html

Code:

• raytracer.cpp - Main Raytracer code, runs independently
• raytracer_post_acceleration.cpp - Early BVH implementation of raytracer (see limtations section below), runs independently

Include Opencv build folder in the Include Directories and Opencv build/x64/vc15/lib folder in the library directories. The following link may help: https://learnopencv.com/code-opencv-in-visual-studio/#Step-4 Also include Eigen in the Additional Include Directories

• Texture maps are only enabled for spheres right now
• Normals are being interpolated on flat surfaces of polygonal meshes, which can cause undesired effects if the flat surface is made reflective
• Only one 3D Procedural texture has been implemented (see the "tex.type ==" if conditions in getPixelColorFromRay)
• Cylinder entity is not fully implemented yet and is buggy
• Bounding Volume Heirarchy is not implemented in the main raytracer program (raytracer.cpp) but instead in raytracer_post_acceleration.cpp which only includes the phong shading model and nonadaptive pixel average supersampling. I had trouble getting my BVH implementation to work well with shadow rays so I moved on without BVH for now.

Inspired by Pixar Inspired by the video game Halo