torchsar.imaging package

Submodules

torchsar.imaging.range_doppler_mftd module

torchsar.imaging.range_doppler_mftd.rda_mftd(Sr, pdict, mfmod='fftconv', iqc=False, rcmc=False, dcem=None, win=None, afa=None, ftshift=False, isplot=False, islog=True)

Range Doppler Algorithm with time domain matched filters

Parameters

  • Sr (Tensor) – The SAR complex echo data representated in real format \({\bm S}_r \in \mathbb{R}^{N_a×N_r×2}\).

  • pdict (dict) – Parameters used in RDA, which are show as follows: {'Vr', 'R0', 'La', 'Fc', 'Tp', 'Kr',  Fsr', 'Fsa'}

  • mfmod (str, optional) –

    Matched filter mode, supported are:

    • 'corr1' : 1d correlation filter and use standard correlation

    • 'conv1' : 1d convolution filter and use standard convolution

    • 'fftcorr1' : 1d correlation filter and use fft for operating correlation

    • 'fftconv1' : 1d convolution filter and use fft for operating convolution (default)

  • iqc (bool, optional) –

    Whether do IQ data correction, see iq_correct():

    • I/Q bias removal

    • I/Q gain imbalance correction

    • I/Q non-orthogonality correction

  • rcmc (bool, int, optional) – Range Migration Correction: integer–>kernel size, False–>no rcmc (default: {8})

  • dcem (str, optional) – Dopplor centroid frequency estimation method (the default is None, does not estimate) - 'abdce_wda' :

  • win (list, tuple or None, optional) – the window function for matched filter of azimuth and range. If None, no window is added (default), e.g. [‘kaiser 12’, ‘hanning’], this will add kaiser window and hanning window in azimuth and range respectively.

  • afa (str, optional) – Dopplor rate estimation (autofocus) method (the default is None, does not do autofocusing)

  • ftshift (bool, optional) – Whether to shift zeros frequency to center when use fft, ifft, fftfreq (the default is False)

  • isplot (bool, optional) – Plot part of processing result, such as DCE result (default: False)

  • islog (bool, optional) – Display processing info (default: True)

Returns

Focused complex image \({\bm S}_i \in \mathbb{C}^{N_a×N_r}\).

Return type

Tensor

Module contents