3DCOMET: 3D Compression Methods Test Dataset

Documentation

Technical details

The dataset contains a total of 51 3D point clouds in PCD (Point Cloud Data) format. As we stated before, we divide our dataset into real and synthetic point clouds. For each type of data (real or synthetic) we then subdivide into three new levels of structure appearance which represent high, medium or low structure presence. Then, for each structure division we subdivide into three levels of texture appearance. On each category we have 5 different point clouds in order to add some variability to the dataset. Finally, we added an special group in order to include more challenging point clouds to compress. These special point clouds are divided into real and synthetic and there are six point clouds on each category.

The real category of the dataset was acquired using the PrimeSense Carmine 1.09a and the Microsoft Kinect 3D sensors. Moreover, some point clouds came from the TUM-RGBD-Dataset (J. Sturm et. al). Each frame in the dataset consists of a point cloud file (pcd) representing the scene.

The synthetic category was generated using our own rendering tool for simulating as if a 3D sensor was placed in the scene (partial 3D view). The capture of synthetic 3D information is made using Java3D. We use different loaders to introduce objects in a scene that are positioned to create scenes. Primitives as cubes, spheres an planes are used too. In our case, we use different scenes to create artificially the different degrees of structure in the dataset. For each created scene we use PickSegment objects to simulate beams according a pinhole camera model, and with PickIntersection and PickResult we get the intersection point coordinates and its rob color. We fixed some restrictions in geometry of the virtual camera to approximate Kinect parameters: horizontal opening 57, vertical 43, resolution of 640x480 pixels and maximum distance. Java3D features allow textures and colors to be applied to objects. Thus we can compose scene variations and evaluate the effect of textures in compression methods.