The RGB-D Scenes Dataset v2 consists of 14 scenes containing furniture (chair, coffee table, sofa, table) and a subset of the objects in the RGB-D Object Dataset (bowls, caps, cereal boxes, coffee mugs, and soda cans). Each scene is a point cloud created by aligning a set of video frames using Patch Volumes Mapping*.
However, the original file is not in TUM RGB-D format, which is not convenient for ORB-SLAM2 to use. This script converts its format.
mkdir build
cd build
cmake ..
make
Organize the original file as follows:
path/
-scene_01 (01 is the sequence number)
-01.pose
then run the ConvertRGBD:
build/ConvertRGBD path 01(number)
The results are as follows:
path/
-scene_01 (01 is the sequence number)
-01.pose
depth/
rgb/
associations.txt
depth.txt
rgb.txt
groundtruth.txt
Since the original file has no explicit timestamp, the converted timestamp is the frame interval (50ms), which can be modified.