highpass_imagehighpass_imageHighpassImageHighpassImage (Operator)


highpass_imagehighpass_imageHighpassImageHighpassImage — Extract high frequency components from an image.


highpass_image(Image : Highpass : Width, Height : )

Herror highpass_image(const Hobject Image, Hobject* Highpass, const Hlong Width, const Hlong Height)

Herror T_highpass_image(const Hobject Image, Hobject* Highpass, const Htuple Width, const Htuple Height)

void HighpassImage(const HObject& Image, HObject* Highpass, const HTuple& Width, const HTuple& Height)

HImage HImage::HighpassImage(Hlong Width, Hlong Height) const

static void HOperatorSet.HighpassImage(HObject image, out HObject highpass, HTuple width, HTuple height)

HImage HImage.HighpassImage(int width, int height)


highpass_imagehighpass_imageHighpassImageHighpassImageHighpassImage extracts high frequency components in an image by applying a linear filter with the following matrix (in case of a 7x5 matrix): 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -35 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 This corresponds to applying a mean operator (mean_imagemean_imageMeanImageMeanImageMeanImage), and then subtracting the original gray value. A value of 128 is added to the result, i.e., zero crossings occur for 128.

This filter emphasizes high frequency components (edges and corners). The cutoff frequency is determined by the size (Height x Width) of the filter matrix: the larger the matrix, the smaller the cutoff frequency is.

At the image borders the pixels' gray values are mirrored. In case of over- or underflow the gray values are clipped (255 and 0, resp.).


If even values are passed for HeightHeightHeightHeightheight or WidthWidthWidthWidthwidth, the operator uses the next larger odd value instead. Thus, the center of the filter mask is always uniquely determined.

highpass_imagehighpass_imageHighpassImageHighpassImageHighpassImage can be executed on OpenCL devices. The same limitations as for mean_imagemean_imageMeanImageMeanImageMeanImage and sub_imagesub_imageSubImageSubImageSubImage apply.

Note that filter operators may return unexpected results if an image with a reduced domain is used as input. Please refer to the chapter Filters.

Execution Information


ImageImageImageImageimage (input_object)  (multichannel-)image(-array) objectHImageHImageHobject (byte* / uint2*) *allowed for compute devices

Input image.

HighpassHighpassHighpassHighpasshighpass (output_object)  (multichannel-)image(-array) objectHImageHImageHobject * (byte / uint2)

High-pass-filtered result image.

WidthWidthWidthWidthwidth (input_control)  extent.x HTupleHTupleHtuple (integer) (int / long) (Hlong) (Hlong)

Width of the filter mask.

Default value: 9

Suggested values: 3, 5, 7, 9, 11, 13, 17, 21, 29, 41, 51, 73, 101

Typical range of values: 3 ≤ Width Width Width Width width ≤ 501

Minimum increment: 2

Recommended increment: 2

Restriction: Width >= 3 && odd(Width)

HeightHeightHeightHeightheight (input_control)  extent.y HTupleHTupleHtuple (integer) (int / long) (Hlong) (Hlong)

Height of the filter mask.

Default value: 9

Suggested values: 3, 5, 7, 9, 11, 13, 17, 21, 29, 41, 51, 73, 101

Typical range of values: 3 ≤ Height Height Height Height height ≤ 501

Minimum increment: 2

Recommended increment: 2

Restriction: Height >= 3 && odd(Height)

Example (C)



highpass_imagehighpass_imageHighpassImageHighpassImageHighpassImage returns 2 (H_MSG_TRUE) if all parameters are correct. If the input is empty the behavior can be set via set_system('no_object_result',<Result>)set_system("no_object_result",<Result>)SetSystem("no_object_result",<Result>)SetSystem("no_object_result",<Result>)SetSystem("no_object_result",<Result>). If necessary, an exception is raised.

Possible Successors

thresholdthresholdThresholdThresholdThreshold, skeletonskeletonSkeletonSkeletonSkeleton


mean_imagemean_imageMeanImageMeanImageMeanImage, sub_imagesub_imageSubImageSubImageSubImage, convol_imageconvol_imageConvolImageConvolImageConvolImage, bandpass_imagebandpass_imageBandpassImageBandpassImageBandpassImage

See also