BazAR is an augmented reality library based on efficient computer vision techniques. It contains an easy to use geometric and photometric camera calibration system. It works in real-time with one or more cameras. A simple OpenGL GLSL shader augments the video stream by inserting artificial 3D objects, shaded with the calibrated lightmap, on a real surface. BazAR is based on a very fast textured planar object detector and features point classification trees.
|Tags||multimedia Graphics Software Development Libraries Scientific/Engineering Image Recognition|
|Operating Systems||Mac OS X Windows POSIX|
Release Notes: A multiple camera calibration bug has been fixed, together with an augmentation accuracy bug that also appeared in the multiple camera case.
Release Notes: Photometric calibration is included. Portable OpenGL relighting with the computed irradiance map.
Release Notes: Geometric single and multi camera calibration is included, with an example showing real-time 3D augmentation.
Release Notes: This initial public release includes a sample application that quickly and automatically detects a sheet of paper on images.