Linear Feature detection

Contents:

• Detection parameters
• Examples
• Table of removestars parameters
• Table of bright parameters
• Table of dim parameters
• Module: removestars
• Module: detecttrails
• Module: processfield

Tip

Description of the algorithm’s implementation, execution parameters and performance is availible in

Bektesevic & Vinkovic, 2017, Linear feature detection algorithm for astronomical surveys - I. Algorithm description

availible at arxiv.org/abs/1612.04748

This module contains all the basic functionality required to execute the linear feature detector interactively or in a sequential batch-like processing setup.

The module requires that all the required data exists and is linked to as described in the setup section.

Central construct of this module is the DetectTrails class which keeps track of the execution parameters and targeted data.

Broadly speaking the detection is a three step process:

• Removal of all known objects
• Detection of bright linear features
• Detection of dim linear features.

Each step is individually configurable through the params dictionaries of the class.

It is instructive to read the docstring of process_field() to see the steps algorithm does. For its implementation or mode details see lfd.detecttrails.processfield module of the package.

The DetectTrails is not SDSS data agnostic but the processing functionality in lfd.detecttrails.processfield is. That way this module can be used to run on other data given that the images are in FITS format and that the catalog for each image is also given as a header table of a FITS file. There should be one such fits catalog per image. Other catalog sources are possible but not impemented. It is instructive to see removestars module which should contain examples of old deprecated CSV catalog functionality.