search

svip-lab / PlanarReconstruction

[CVPR'19] Single-Image Piece-wise Planar 3D Reconstruction via Associative Embedding
MIT License
358 stars 85 forks source link

Custom dataset #41

Open KirillHiddleston opened 3 years ago

KirillHiddleston commented 3 years ago

explain to me what num_planes is, where can I get it?

niujinshuchong commented 3 years ago

@KirillHiddleston Please refer to https://github.com/svip-lab/PlanarReconstruction/blob/master/main.py#L107 to prepare custom dataset.

KirillHiddleston commented 3 years ago

sorry explain to me what "num_planes" is, where can I get it?

niujinshuchong commented 3 years ago

@KirillHiddleston num_planes is the the number of plane instances in the image.

KirillHiddleston commented 3 years ago

shape set of image load in your npz file

image (192, 256, 3) depth (192, 256, 1) segmentation (192, 256, 1) index map plane (20, 3) ?

how to get plane ? I have normal map can u explain ?

https://github.com/art-programmer/PlaneNet A plane is represented by three parameters and a segmentation mask. If the plane equation is nx=d where n is the surface normal and d is the plane offset, then plane parameters are nd. The plane equation is in the camera frame, where x points to the right, y points to the front, and z points to the up.

niujinshuchong commented 3 years ago

@KirillHiddleston If you have segmentation map of plane and depth map. Then you can unproject depth map to point cloud, then you can fit the plane parameter to 3D points of this plane.