This package holds header and format definitions for CSV-files that hold timestamped 3D spatial information. By spatial

• 3-DoF relative position (),
• 3-DoF attitude ,
• 6-DoF pose (position + orientation represented by quaternions in the [qx qy qz qw] order)
• 3-DoF position and 3-DoF orientation with uncertainty (position and orientation uncertainties is given by two 3x3 upper triangular covariance matrices). TODO: make default assumption on the uncertainty space
• 3-DoF position and 3-DoF orientation with typed uncertainty (position and orientation uncertainties is given by two 3x3 upper triangular covariance matrices) [1*].
• 6-DoF pose with upper triangular pose uncertainty. TODO: make default assumption on the uncertainty space
• 6-DoF pose with upper triangular pose typed uncertainty [1*] .

[1*] Typed uncertainty: The space/reference frame of the covariance is specified by the "est_err_type" and "err_representation" entry. In case of an error-state estimator, the error representation of the orientation needs to be specified in the CSV files. The estimation error types are defined in the EstimationErrorType file. The error representation type is defined in the ErrorRepresentationType file.

Orientation are represented by quaternions in the [qx qy qz qw] order, meaning that the real-part appears at the end aka JPL order.

File headers are in the first line of a CSV file should not start with a #, followed by a sequence of unique comma separated strings/chars.

It is highly recommended loading the CSV files into a pandas.DataFrame. For convenience, there is a package called cnspy_csv2dataframe that does the conversion using the CSVFormatPose definitions.

The CSVFormatPose.TUM format, got it's name for file format used in the TUM RGB-D benchmark tool. Noticeable, is that the order of quaternion is non-alphabetically ([q_x,q_y,q_z, q_w] instead of [q_w, q_x, q_y, q_z]), meaning that first comes the imaginary part, then the real part, but this is just a matter of taste and definition! To be backward compatible with older/other tools (TUM RGB-D benchmark tool, rpg_trajectory_evaluation, etc.), we follow this non-alphabetically order! Note that the rpg_trajectory_evaluation framework is based on space-separated *.txt trajectory files, meaning that these files cannot be directly processed in the current framework as the file header cannot be interpreted correctly. Support may be added in future.

Install the current code base from GitHub and pip install a link to that cloned copy

git clone https://github.com/aau-cns/spatial_csv_formats.git
cd spatial_csv_formats
pip install -e .

or the official package via

pip install cnspy-spatial-csv-formats

It is part of the cnspy eco-system of the cns-github group.

• enum

Software License Agreement (GNU GPLv3 License), refer to the LICENSE file.

Sharing is caring! - Roland Jung