torchbox.imaging package

Submodules

torchbox.imaging.backprojection module

torchbox.imaging.backprojection.bp_ct()
torchbox.imaging.backprojection.bp_sar(Src, traj, grid, dr, fc, dcpmode=1, ftshift=False, batchsize=128)

FMCW SAR backprojection imaging

Parameters:

  • Src (tensor) – range compressed signal, .. math:: {mathbb R}^{N_a time N_r}

  • traj (tensor) – radar trajectory .. math:: {mathbb R}^{N_a}

  • grid (tensor) – imaging grids at x, y, z axis, with size (..., 3)

  • dr (float) – range cell resolution

  • fc (float) – carrier frequency

  • dcpmode (int, optional) – dechirp mode, by default 1

  • ftshift (bool, optional) – fft shifting flag, by default False

  • batchsize (int, optional) – processing batch size in azimuth, by default 128

Returns:

imaging results

Return type:

tensor

torchbox.imaging.backprojection.bp_sar_fmcw(Sif, traj, grid, dr, fc, ups=None, dcpmode=1, ftshift=False, batchsize=128)

FMCW SAR backprojection imaging

Parameters:

  • Sif (tensor) – intermediate frequency signal, .. math:: {mathbb R}^{N_a time N_r}

  • traj (tensor) – radar trajectory .. math:: {mathbb R}^{N_a}

  • grid (tensor) – imaging grids at x, y, z axis, with size (..., 3)

  • dr (float) – range cell resolution

  • fc (float) – carrier frequency

  • ups (str or None, optional) – 'fftA' upsampled by fft with upsampling rate A, by default None (linear interpolation)

  • dcpmode (int, optional) – dechirp mode, by default 1

  • ftshift (bool, optional) – fft shifting flag, by default False

  • batchsize (int, optional) – processing batch size in azimuth, by default 128

Returns:

imaging results

Return type:

tensor

torchbox.imaging.backprojection.imaging_grid(xgi=None, ygi=None, zgi=None, order='xy', device='cpu')

generates grids for imaging

Parameters:

  • xgi (list, tuple or None, optional) – x grid information, [xmin, xmax, Nx], by default None (not generated)

  • ygi (list, tuple or None, optional) – y grid information, [ymin, ymax, Ny], by default None (not generated)

  • zgi (list, tuple or None, optional) – z grid information, [zmin, zmax, Nz], by default None (not generated)

  • order (str, optional) – output order, by default 'xy'

  • device (str, optional) – device string, by default 'cpu'

Returns:

imaging grids: gx, gy, gz

Return type:

tuple

Raises:

ValueError – Only support 3D grid, i.e. ‘’x’`, 'y', 'z' in order.

Examples

grid = imaging_grid(xgi=[-3, 3, 10], ygi=[0, 5, 5], zgi=None, order='yx', device='cpu')
print(grid[0].shape, grid[1].shape, grid[0].device, grid[1].device)

grid = imaging_grid(xgi=[-3, 3, 10], ygi=[0, 5, 5], zgi=[-2, 2, 6], order='xyz', device='cpu')
print(grid[0].shape, grid[1].shape, grid[2].shape, grid[0].device, grid[1].device, grid[2].device)

# output
torch.Size([10, 5]) torch.Size([10, 5]) cpu cpu
torch.Size([5, 10, 6]) torch.Size([5, 10, 6]) torch.Size([5, 10, 6]) cpu cpu cpu

Module contents