Tcanny

Revision as of 03:24, 16 February 2016 (view source) Reel.Deal (Talk | contribs) m (remove space) ← Older edit		Revision as of 02:08, 11 June 2016 (view source) Reel.Deal (Talk | contribs) m (link) Newer edit →
Line 3:		Line 3:
	| {{Author/tritical}}		| {{Author/tritical}}
	| v1.0		| v1.0
−	|[http://bengal.missouri.edu/~kes25c/tcanny.zip tcanny.zip]	+	|[http://web.archive.org/web/20140804144614/http://bengal.missouri.edu/~kes25c/tcanny.zip tcanny.zip]
	| Edge Detection		| Edge Detection
	| [http://www.gnu.org/licenses/gpl-2.0.txt GPLv2]		| [http://www.gnu.org/licenses/gpl-2.0.txt GPLv2]

---

Revision as of 02:08, 11 June 2016

Contents 1 Description 2 Requirements 3 Syntax and Parameters 3.1 tcanny 3.2 tdtrans 4 Examples 5 Changelog 6 Archived Downloads 7 External Links

Description

tcanny contains a Canny edge detection filter and distance transform filter.

Requirements

• AviSynth 2.5.8 or greater
• Supported color formats: YUY2, YV12

Syntax and Parameters

tcanny

Builds an edge map using Canny edge detection.

tcanny(clip, float "sigma", float "t_h", float "t_l", int "nms", int "mode", int "plane", float "gmmax")

clip   =

Input clip; only YUY2 and YV12 are supported.

float  sigma = 1.5

Standard deviation of Gaussian blur. 0 means no blurring before edge detection.

float  t_h = 8.0

High gradient magnitude threshold for hysteresis.

float  t_l = 1.0

Low gradient magnitude threshold for hysteresis.

int  nms = 3

Byte mask choosing non-maxima suppression method(s).

If (nms&1) - if pixel is maxima along nearest integer direction then keep.

If (nms&3) - if pixel is maxima along linearly interpolated direction then keep.

int  mode = 0

Sets the output format:

• 0 : thresholded edge map (255 for edge, 0 for non-edge)
• 1 : gradient magnitude map. Scaled to 0-255 range based on gmmax.

pixel_val = min(grad_mag*255/gmmax,255)

• 2 : edge pixel only gradient direction map (non-edge pixels set to 0)
• 3 : gradient direction map

            Gradient direction maps discretize [0-180) degree range into 256 bins.

                 0 = [000.000-000.351}
                 1 = [000.351-001.054)
                 .
                 .
                 .
               255 = [179.648-180.000)

                        135   90  45
                           \  |  /
                       180 --   -- 0
                            / |  \
                          45  90  135

                * This is the gradient direction, so the line direction 
                  is orthogonal to it (90 degree rotation). It only goes 
                  through 180 because for this case:

                     180+n = n, 0 <= n < 180

int  plane = 1

Byte mask indicating which planes to process. If a plane isn't processed then it is copied through as is.

if (plane&1) - process Y

if (plane&2) - process U

if (plane&4) - process V

float  gmmax = 50.0

Used for scaling gradient magnitude into [0,255] for mode=1.

gmmax is internally set to 1.0 if you set it to < 1.0.

---

tdtrans

Performs a distance transformation on an edge map. This filter assumes the input image is 0 for non-edge pixels, and > 0 for edge pixels.
The distance transform sets each pixel equal to the distance from it to the nearest edge pixel.

tdtrans (clip, int "mode", int "plane")

clip   =

Input clip; only YUY2 and YV12 are supported.

int  mode = 1

Sets distance type.

• 0 - city block
• 1 - euclidean

* For the euclidean case the distance is rounded to the nearest integer.

* For both cases distances greater than 255 are set to 255.

int  plane = 1

Byte mask indicating which planes to process. Same as in tcanny.

Examples

tcanny with default settings:

AviSource("blah.avi")
TCannyMod(sigma=1.5, t_h=8.0, t_l=1.0, nms=3, mode=0, plane=1, gmmax=50.0)

tdtrans with default settings:

AviSource("blah.avi")
tdtrans(mode=1, plane=1)

Changelog

Version      Date            Changes

v1.0         06/03/2009      - Initial release

Archived Downloads

Version	Download	Mirror
v1.0	tcanny.zip	tcanny.zip

External Links

---

Back to External Filters ←