With Visual effects becoming an increasingly popular mode of storytelling, product showcasing, and artistic expression, camera tracking has become a more commonly faced hurdle. At the very root of any visual effects project is the process of camera tracking. Visual effects (VFX) is the process of integrating computer generated visuals into real-life camera footage.

Oftentimes, these computer generated images come from 3D models. To composite these two media sources, the 3D model has to move with the real environment in the camera’s footage, but this is only possible if the motion of the camera is known. Fortunately, the camera’s motion can be reconstructed with computer vision algorithms that look at the footage. While the algorithms under the hood do much of the heavy lifting for an artist, the process of camera tracking is still monotonous. There is a learning curve to get reliable results and even for experienced artists, the process can be time consuming.

The goal is to provide a minimal-effort solution for solo artists that is fairly robust and accurate as well. VFX is time consuming and therefore a significant business expense. The ending product will significantly cut down on the time required for the camera tracking step by completely automating the process of finding high-contrast points, culling high error tracks, and more.

Photogrammetry Add-On Panel Inside of Blender's UI

An early implementation of our Camera Tracking algorithm.

sddec25 poster

A demo of the working Add-On.