WISE: Wavelet Image Segmentation and Evaluation

The WISE package has been developed to address the issue of detecting significant features in radio interferometric images and obtaining reliable velocity field from cross-correlation of these regions in multi-epoch observations.

It comprises three main constituents:

  • Detection of structural information is performed using the segmented wavelet decomposition (SWD) method. The SWD algorithm provides a structural representation of astronomical images with exceptional sensitivity for identifying compact and marginally resolved features as well as large scale structural patterns. It delivers a set of two-dimensional significant structural patterns (SSP), which are identified at each scale of the wavelet decomposition.
  • Tracking of these SSP detected in multiple-epoch images is performed with a multiscale cross-correlation (MCC) algorithm. It combines structural information on different scales of the wavelet decomposition and provides a robust and reliable cross-identification of related SSP.
  • A stacked cross correlation (SCC) is introduced to recover multiple velocity components from partially overlapping emitting regions.
_images/m87_combined.png

This software is based on a method introduced and described in:

“Wavelet-based decomposition and analysis of structural patterns in astronomical image”
Mertens, F., Lobanov, A.P. 2015, Astronomy & Astrophysics, 574, 67 ADS
“Detection of multiple velocity components in partially overlapping emitting regions”
Mertens, F., Lobanov, A.P. 2016, Astronomy & Astrophysics, 587, 52 ADS

Installation

The WISE package can be installed using the conda package management system. You can install conda easily by downloading the installer corresponding to your OS and for Python 2.7 at:

http://conda.pydata.org/miniconda.html

and following the installation instruction at:

http://conda.pydata.org/docs/install/quick.html

Once done, you can install WISE in just one command line:

conda install -c flomertens -c conda-forge wise cython dask

It is also possible to install WISE using pip:

pip install wisetool

Alternative manual installation:

Latest version of WISE and libwise package can be download at:

https://github.com/flomertens/libwise/releases/

https://github.com/flomertens/wise/releases/

Each package can be installed with the command line:

python setup.py install

Requirements:

Using WISE

WISE can be used either from the command line or from an ipython notebook. Check the tutorials below to see how to use it.

Indices and tables