A ROS node for basic morphologic track classification.
To run the Python code, install the virtual environment by running:
./create_python_env.shCoordinates of each cluster are relative to its top-left corner, the coordinates within the cluster are then relative to its position.
roslaunch classification real_time.launchThe rospix output:
/track_classifier/image_in
The ROS image (possible to view using, e.g., rqt_image_view) showing labeled clusters:
/track_classifier/labeled_image
The classified data per each image:
/track_classifier/data
In the batch mode, the ROS node takes CSV files in a given folder and processed them sequentially.
The node expects 1 image per input file, and the filenames in the format filenamename.extension.
As results, it outputs two files per image, the 1st one containing overall image statistics, the 2nd one containing a cluster list. Coordinates of each cluster are relative to its top-left corner, the coordinates within the cluster are then relative to its position. The results, containing the statistics for each image and a list of clusters, are saved in JSON format.
The input and the output folders are set in the config file config/batch_processing.yaml.
# path will be relative to this package root folder
path_relative: True # use False if you plan on providing absolute filepaths
source_path: "datasets/in"
result_path: "datasets/out"
classifier_pipeline_path: "pipelines/vzlusat-1.joblib"
# delimiter for data on the input
input_delimiter: ','To run the batch processing, issue the command:
roslaunch classification batch.launch- T. Baca, D. Turecek, R. McEntaffer and R. Filgas, Rospix: Modular Software Tool for Automated Data Acquisitions of Timepix Detectors on Robot Operating System, Journal of Instrumentation 13(11):C11008, 2018.
- M. Jilek, Processing of Radiation Data from the Timepix Sensor on the VZLUSAT-1 Satellite, Master's thesis, Czech Technical University in Prague, Faculty of Electrical Engineering, 2018.
- T. Baca, M. Jilek, I. Vertat, M. Urban, O. Nentvich, R. Filgas, C. Granja, A. Inneman and V. Daniel Timepix in LEO Orbit onboard the VZLUSAT-1 Nanosatellite: 1-year of Space Radiation Dosimetry Measurements, Journal of Instrumentation 13(11):C11010, 2018.
The work has been done on behalf of Medipix2 collaboration and is currently supported by the Czech Science Foundation project 18-10088Y and by Czech Technical University grant no. SGS17/187/OHK3/3T/13.