tfio.experimental.filter.gabor

Apply Gabor filter to image.

tfio.experimental.filter.gabor(
    input, freq, sigma=None, theta=0, nstds=3, offset=0, mode=None,
    constant_values=None, name=None
)

input A 4-D ([N, H, W, C]) Tensor.
freq A float Tensor. Spatial frequency of the harmonic function. Specified in pixels.
sigma A scalar or 1-D [sx, sy] Tensor. Standard deviation in in x- and y-directions. These directions apply to the kernel before rotation. If theta = pi/2, then the kernel is rotated 90 degrees so that sigma_x controls the vertical direction. If scalar, then sigma will be broadcasted to 1-D [sx, sy].
nstd A scalar Tensor. The linear size of the kernel is nstds standard deviations, 3 by default.
offset A scalar Tensor. Phase offset of harmonic function in radians.
mode A string. One of "CONSTANT", "REFLECT", or "SYMMETRIC" (case-insensitive). Default "CONSTANT".
constant_values A scalar, the pad value to use in "CONSTANT" padding mode. Must be same type as input. Default 0.
name A name for the operation (optional).

A 4-D ([N, H, W, C]) Tensor.