is an option to various array and image operations that specifies what padding to use when extending beyond the original data specified.


  • Padding is typically used in filtering, resizing or spatial transformations, whenever the operation extends beyond the original data domain.
  • Typical settings for Padding include:
  • valpad with a constant value val
    "Fixed"repetitions of the elements on each boundary
    "Periodic"cyclic repetitions of the complete array
    "Reflected"reflections of the array in the boundary
    "Reversed"reversals of the complete array
  • For a 2D array, "Fixed" indicates that the rectangle of elements added at each corner should be copies of the elements at the corners of the original array.
  • "Reversed" indicates that the outermost layer of elements in the array should be repeated as the innermost layer of elements in the padding. "Reflected" specifies that these elements should not be repeated.


open allclose all

Basic Examples  (2)

Specify a padding when resampling a list:

Use a different scheme:

Image convolution with a Gaussian, taking pixel values outside the image to be yellow:

Scope  (1)

Many filter functions use Padding->"Fixed" to minimize visibility of padding artifacts:

Blur an example image using default padding method:

Use zero padding:

Wolfram Research (2008), Padding, Wolfram Language function,


Wolfram Research (2008), Padding, Wolfram Language function,


@misc{reference.wolfram_2020_padding, author="Wolfram Research", title="{Padding}", year="2008", howpublished="\url{}", note=[Accessed: 26-January-2021 ]}


@online{reference.wolfram_2020_padding, organization={Wolfram Research}, title={Padding}, year={2008}, url={}, note=[Accessed: 26-January-2021 ]}


Wolfram Language. 2008. "Padding." Wolfram Language & System Documentation Center. Wolfram Research.


Wolfram Language. (2008). Padding. Wolfram Language & System Documentation Center. Retrieved from