Wolfram Language & System 10.3 (2015)|Legacy Documentation

This is documentation for an earlier version of the Wolfram Language.View current documentation (Version 11.2)


gives an image corresponding to the magnitude of the gradient of image, computed using discrete derivatives of a Gaussian of pixel radius r.

uses a Gaussian with standard deviation σ.

uses a Gaussian with radii etc. in the vertical and horizontal directions.

applies gradient filtering to an array of data.

Details and OptionsDetails and Options

  • GradientFilter works with arbitrary 2D and 3D images, as well as data arrays of any rank.
  • For a single-channel image and for data, the gradient magnitude is the Euclidean norm of the gradient at a pixel position, approximated using discrete derivatives of Gaussians in each dimension.
  • For multichannel images, define the Jacobian matrix to be , where is the gradient for channel . The gradient magnitude is the square root of the largest eigenvalue of .
  • GradientFilter[image,] always returns a single-channel image.
  • GradientFilter[image,r] is equivalent to GradientFilter[image,{r,r/2}].
  • The following options can be specified:
  • MethodAutomaticconvolution kernel
    Padding"Fixed"padding method
    WorkingPrecisionAutomaticthe precision to use
  • The following suboptions can be given to Method:
  • "DerivativeKernel""Bessel"convolution kernel
    "NonMaxSuppression"Falsewhether to use non-maximum suppression
  • Possible settings for include:
  • "Bessel"standardized Bessel derivative kernel, used for Canny edge detection
    "Gaussian"standardized Gaussian derivative kernel, used for Canny edge detection
    "ShenCastan"first-order derivatives of exponentials
    "Sobel"binomial generalizations of the Sobel edge-detection kernels
    {kernel1,kernel2,}explicit kernels specified for each dimension
  • GradientFilter[image,] by default gives an image of the same dimensions as image.
  • With a setting Padding->None, GradientFilter[image,] normally gives an image smaller than image.
Introduced in 2008
| Updated in 2015