Implementations of feature detection methods.


Interface for the feature detection module. It returns a callable function for detecting features from two-dimensional images.

The feature detectors implement the following interface:

detection(input_image, **keywords)

The input is a two-dimensional image. Additional arguments to the specific method can be given via **keywords. The output is an array of shape (n, m), where each row corresponds to one of the n features. The first two columns contain the coordinates (x, y) of the features, and additional information can be specified in the remaining columns.

All implemented methods support the following keyword arguments:




maximum number of features to detect


Return a callable function for feature detection.


Blob detection methods.

detection(input_image[, max_num_features, ...])


Thunderstorm cell detection module, part of Thunderstorm Detection and Tracking (DATing) This module was implemented following the procedures used in the TRT Thunderstorms Radar Tracking algorithm ([HMG+04]) used operationally at MeteoSwiss. Full documentation is published in [FGGB21]. Modifications include advecting the identified thunderstorms with the optical flow obtained from pysteps, as well as additional options in the thresholding.


[HMG+04] [FGGB21]

@author: mfeldman

detection(input_image[, max_num_features, ...])

This function detects thunderstorms using a multi-threshold approach.

breakup(ref, minval, maxima)

This function segments the entire 2-D array into areas belonging to each identified maximum according to a watershed algorithm.

longdistance(loc_max, mindis)

This function computes the distance between all maxima and rejects maxima that are less than a minimum distance apart.

get_profile(areas, binary, ref, loc_max, ...)

This function returns the identified cells in a dataframe including their x,y locations, location of their maxima, maximum reflectivity and contours.


Shi-Tomasi features detection method to detect corners in an image.

detection(input_image[, max_corners, ...])