search

google-research-datasets / Objectron

Objectron is a dataset of short, object-centric video clips. In addition, the videos also contain AR session metadata including camera poses, sparse point-clouds and planes. In each video, the camera moves around and above the object and captures it from different views. Each object is annotated with a 3D bounding box. The 3D bounding box describes the object’s position, orientation, and dimensions. The dataset contains about 15K annotated video clips and 4M annotated images in the following categories: bikes, books, bottles, cameras, cereal boxes, chairs, cups, laptops, and shoes
Other
2.24k stars 263 forks source link

Near and far plane? #27

Open deeplearningfam opened 3 years ago

deeplearningfam commented 3 years ago

Hi,

thanks a lot for the great dataset! I was wondering if it is possible to get the approximate near and far plane for the frames? I couldn't find this in the geometry tutorial.

Thanks!

ahmadyan commented 3 years ago

The near and far plane are optimized per frame, and they are available via the projection_matrix, specifically, proj[2, 2] and proj[2, 3]

OpenGL projection matrix )

Which after solving should give you 1000m and 1cm, but depending on calibration it might vary.