I'm working on depth estimation using the NYU Depth Dataset V2. I have an RGB image and its corresponding Kinect depth map which is in black and white. I need to convert the black and white depth map to an RGB depth map. What is the best way to proceed with this? The documentation does not provide enough information about the procedure. I'm attaching the RGB images and their b/w kinetic depth maps from the matlab file of the dataset. Sorry for the bad quality of the images.
Else, what other datasets can be used for depth estimation to create depth maps using RGB images.
From your question I assume that you would like to map the depth map to the RGB image, so that you have the depth for every pixel in the RGB image.
The same dataset includes a set of Matlab scripts, one of them, named project_depth_map.m, can be used to do the mapping You can download the toolbox here: http://cs.nyu.edu/~silberman/code/toolbox_nyu_depth_v2.zip
To do this mapping one needs to know the parameters of both cameras, RGB and IR. With the parameters you can have the coordinates from a pixel in one image to the other.
From your question I assume that you would like to map the depth map to the RGB image, so that you have the depth for every pixel in the RGB image.
The same dataset includes a set of Matlab scripts, one of them, named project_depth_map.m, can be used to do the mapping You can download the toolbox here: http://cs.nyu.edu/~silberman/code/toolbox_nyu_depth_v2.zip
To do this mapping one needs to know the parameters of both cameras, RGB and IR. With the parameters you can have the coordinates from a pixel in one image to the other.
- Badges
- Science topic
- Similar topics
- Filing