deltametrics.mobility.calculate_channelized_response_variance

deltametrics.mobility.calculate_channelized_response_variance(arr, threshold=0.2, normalize_input=False, normalize_output=False)

Calculate the Channelized Response Variance (CRV).

This function takes a t-x-y array and calculates its directional CRV [1]. In short, the function does the following:

1. Normalizes the array at each time slice if desired.

2. Calculates the CRV magnitude (aka variance along time-axis) and normalizes this array if desired.

3. Does linear regressions through time for each pixel and returns the slopes.

4. Calculates the directional CRV using a slope threshold value.

5. Returns the CRV magnitude, slopes, and directional CRV values

[1]

Jarriel, Teresa, Leo F. Isikdogan, Alan Bovik, and Paola Passalacqua. “Characterization of deltaic channel morphodynamics from imagery time series using the channelized response variance.” Journal of Geophysical Research: Earth Surface 124, no. 12 (2019): 3022-3042.

Parameters:

• arr (numpy.ndarray) – A t-x-y 3-D array to calculate the CRV on.

• threshold (float, optional) – Threshold for CRV calculation. The default is 0.2.

• normalize_input (bool, optional) – Whether to normalize the input images pixel values to 0-1. The default is False.

• normalize_output (bool, optional) – Whether to normalize the output image pixel values to 0-1. The default is False.

Returns:

• crv_magnitude (numpy.ndarray) – A t-x-y 3-D array with the CRV magnitude.

• slopes (numpy.ndarray) – A t-x-y 3-D array with the slopes of the linear regressions.

• directional_crv (numpy.ndarray) – A t-x-y 3-D array with the directional CRV.

Examples

(Source code, png, hires.png)