title: EdgeDetectCImg
media types:
Video experimental
description: Perform edge detection by computing the image gradient magnitude. Optionally, edge detection can be preceded by blurring, and followed by erosion and thresholding. In most cases, EdgeDetect is followed a Grade node to extract the proper edges and generate a mask from these.
For color or multi-channel images, several edge detection algorithms are proposed to combine the gradients computed in each channel:
References:
CImg is a free, open-source library distributed under the CeCILL-C (close to the GNU LGPL) or CeCILL (compatible with the GNU GPL) licenses. It can be used in commercial applications (see http://cimg.eu).
version: 2
creator: mr.fantastic mrfantastic@firemail.cc
license: GPLv2
URL: https://openeffects.org/
title: R
description:
Process red component.
type: boolean
readonly: no
required: no
animation: yes
default: 1
title: G
description:
Process green component.
type: boolean
readonly: no
required: no
animation: yes
default: 1
title: B
description:
Process blue component.
type: boolean
readonly: no
required: no
animation: yes
default: 1
title: A
description:
Process alpha component.
type: boolean
readonly: no
required: no
animation: yes
default: 0
title: Filter
description:
Edge detection filter. If the blur size is not zero, it is used as the kernel size for quasi-Gaussian, Gaussian, box, triangle and quadratic filters. For the simple, rotation-invariant and Sobel filters, the image is pre-blurred with a Gaussian filter.
type: string
readonly: no
required: no
animation: yes
default: Gaussian
values:
title: Multi-Channel
description:
Operation used to combine multi-channel (e.g. color) gradients into an edge detector. This parameter has no effect if a single channel (e.g. alpha) is processed.
type: string
readonly: no
required: no
animation: yes
default: Tensor
values:
title: Blur Size
description:
Size of the blur kernel applied before edge detection.
type: float
readonly: no
required: no
animation: yes
minimum: 0
maximum: 100
default: 0
title: Erode Size
description:
Size of the erosion performed after edge detection.
type: float
readonly: no
required: no
animation: yes
minimum: -10
maximum: 10
default: 0
title: Non-Maxima Suppression
description:
Perform non-maxima suppression (after edge detection and erosion): only values that are maximal in the direction orthogonal to the contour are kept. For multi-channel images, the contour direction estimation depends on the multi-channel operation.
type: boolean
readonly: no
required: no
animation: yes
title: Expand RoD
description:
Expand the source region of definition by 1.5size (3.6sigma).
type: boolean
readonly: no
required: no
animation: yes
default: 1
title: (Un)premult
description:
Divide the image by the alpha channel before processing, and re-multiply it afterwards. Use if the input images are premultiplied.
type: boolean
readonly: no
required: no
animation: yes
title: Invert Mask
description:
When checked, the effect is fully applied where the mask is 0.
type: boolean
readonly: no
required: no
animation: yes
title: Mix
description:
Mix factor between the original and the transformed image.
type: float
readonly: no
required: no
animation: yes
minimum: 0
maximum: 1
default: 1
title: Top-Left Origin
description:
Set to 1 to use MLT top-left image origin instead of the OFX bottom-left origin. Use for plugins that crash or produce incorrect output with negative row bytes.
type: boolean
readonly: no
required: no
minimum: 0
maximum: 1
default: 0
widget: checkbox
Subscribe to News via RSS.