CleanTiPy 0.6.2

An open-source implementation of CLEAN-T acoustic imaging technique

CleanTiPy is a python package allowing users to realise acoustic images of moving sources. It implements the CLEAN-T algorithm, with some additional features such as:

• Angular selection of the trajectory (to realise the analysis on portions of the trajectory)
• Frequency band filtering (on third-octave or octave band) to do the analysis for specific frequency band (can speed up the computation in low frequency by decimating the signal)
• Implementation of (an homogeneous) wind effect on propagation and back-propagation.

As implementing CLEAN-T, this package can thus be used for computing classical Beamforming computation and propagation simulations (being constitutive blocks of iterative CLEAN-T algorithm).

It is based on the work published in Cousson et al. and in Leiba et al.

Installation

This code is developed in Python 3.11 and therefore back-compatibility is not guaranteed.

Install the required packages with pip:

pip install cleantipy

Usage

Examples can be found in the examples directory on github

An exemple (CLEAN-T over trajectory, for multiple frequency bands) can be run this way:

cd ./examples/
python computeCleanT_multiFreq.py

An exemple of CLEAN-T over trajectory, for multiple frequency bands, and for multiple angular windows can be run this way:

cd ./examples/
python computeCleanT_multiFreq_multiAngles.py

Same example can be found as notebook here

Documentation

The full documentation of the project is available in pdf here. It can also be built locally using

sphinx-build -b html docs/source/ docs/build/html

Support

Contact Raphaël LEIBA : [e-mail]

Contributing

Not open for contribution yet

Authors and acknowledgment

Raphaël Leiba^1,2, with the help of Quentin Leclère²

¹Joint Research Unit in Environmental Acoustics UMRAE, Gustave Eiffel University
²Laboratory of Vibration and Acoustics LVA, INSA Lyon

License

CleanTiPy is licensed under the EUPL-1.2. See LICENSE

Project status

Ready for production realse but still in active development

Packaging

Reminder: to make the code a python package use :

py -m build