cupyx.scipy.signal.sepfir2d

cupyx.scipy.signal.sepfir2d(input, hrow, hcol)[source]

Convolve with a 2-D separable FIR filter.

Convolve the rank-2 input array with the separable filter defined by the rank-1 arrays hrow, and hcol. Mirror symmetric boundary conditions are assumed. This function can be used to find an image given its B-spline representation.

The arguments hrow and hcol must be 1-dimensional and of off length.

Parameters
  • input (cupy.ndarray) – The input signal

  • hrow (cupy.ndarray) – Row direction filter

  • hcol (cupy.ndarray) – Column direction filter

Returns

The filtered signal

Return type

cupy.ndarray