3D Reconstructions Dataset

Isabel Patiño

We provide 3D reconstructions obtained with KinFu (open-source implementation of KinectFusion available in PCL) of four different objects: a dummy head (35.0 cm height), a soccer ball (11.0 cm radius), an American football (12.7×22.2×12.7 cm), and a rubber duck (23.0×21.0×28.0 cm). The reconstructions (mesh and point cloud) were generated offline with input point clouds obtained, from different distances, with six out-of-the-shelf RGBD sensors: a Microsoft Kinect Xbox 360, a Microsoft Kinect in near mode, an Orbbec Astra S, an Asus Xtion Pro Live, a Microsoft Kinect v2, and an Intel RealSense R200.

The data acquisition process and evaluation of this dataset can be found in:
Patino Mejia, I., and Zell, 3D Reconstructions with KinFu Using Different RGBD Sensors. 3rd IEEE International Conference on Image Processing, Applications and Systems (IPAS 2018). 2018.
 
File Structure
The reconstructions are categorized according to the sensor used to record their input point clouds. The file structure is as follows
    Sensor folder (e.g., Microsoft_Kinect_near_mode):

  • camera_info: camera parameters obtained from the topic published by ROS.
  • KinFu_Output_1revolution_inputPointClouds: containes the  reconstructions generated with all the point clouds obtained in approximately 1 revolution.
    • KinFu_Output_Pmatrix_8fps: containes the reconstructions generated at 8 fps.
    • KinFu_Output_Pmatrix_30fps: containes the reconstructions generated at 30 fps.
  • KinFu_Output_800_inputPointClouds: containes the  reconstructions generated with 800 point clouds distribuited on 1 revolution.
    • KinFu_Output_Pmatrix_8fps: containes the reconstructions generated at 8 fps.
    • KinFu_Output_Pmatrix_30fps: containes the reconstructions generated at 30 fps.

The ground truth meshes obtained with the Artec EVA scanner are also provide as .ply  and .bin  (CloudCompare file) files. Additionally, point clouds after outliers removal are also supplied, they are identified as objectXX_X_cloud2.ply. In case you require the .bag files, please get in contact with Isabel Patiño.

 

Software Version

ROS Indigo

ROS Kinect