GenericFilters [Now part of VS core]

[Chikuzen]

GenericFilters-0.7.0.zip (http://www.mediafire.com/download/pxqrjpq1800lmd5/GenericFilters-0.7.0.zip)

https://github.com/chikuzen/GenericFilters

GenericFilters is a set of common image-processing filters.

I think that the necessity of using masktools2 of avisynth will decrease considerably if this plugin and std.functions are used together.


Cross reference list of masktools2 and GenericFilters

mt_edge - generic.Convolution/generic.Sobel/generic.Prewitt/generic.TEdge
mt_invert - generic.Invert
mt_binarize - generic.Binarize
mt_expand - generic.Maximum
mt_inpand - generic.Minimum
mt_inflate - generic.Inflate
mt_deflate - generic.Deflate
mt_clamp - generic.Limiter
mt_convolution - generic.Convolution/generic.ConvolutionHV
mt_hysteresis - generic.Hysteresis
mt_lut - std.Lut/std.Expr
mt_lutxy -std.Lut2/std.Expr
mt_lutxyz - std.Expr
mt_merge - std.MaskedMerge

EDIT:
20130812 update to 0.7.0

[Myrsloik]

It is my duty to shamelessly bump this post since it renders masktools almost completely obsolete. We're now one big step closer to getting completely native vapoursynth scripts for useful things.

[Daemon404]

It is my duty to shamelessly bump this post since it renders masktools almost completely obsolete. We're now one big step closer to getting completely native vapoursynth scripts for useful things.

I dream of a native OS X / Linux QTGMC.

[sl1pkn07]

https://github.com/chikuzen/GenericFilters/issues/1

[06_taro]

Another thing is, I use the following script to compare the output of mt_edge and generic.Convolution.

clip = something
c1 = core.avs.mt_edge(clip, '1 2 1 0 0 0 -1 -2 -1 1', 0, 255)
c2 = core.generic.Convolution(clip, [1, 2, 1, 0, 0, 0, -1, -2, -1], planes=0)
clip = core.std.Interleave([c1, c2])
clip = core.std.ShufflePlanes(clip, planes=0, format=vs.GRAY)

I originally think the result of the two will be identical, but it's not. However I am no expert at matrix convolution so I may misunderstand its functionality and use it the wrong way. :o

I believe you intended to compare between avs.mt_convolution(clip, '1 2 1', '1 0 -1') and generic.Convolution(clip, [1, 2, 1, 0, 0, 0, -1, -2, -1], planes=0), which looks quite identical to me....

[Chikuzen]

hmm...

generic.Convolution always saturate the output on 8bit format.
mt_edge("1 2 1 0 0 0 -1 -2 -1 1", 0, 255) is same as mt_convolution("1 2 1", "1 0 -1", saturate=false)

currently, I'm working on imprementing Sobel and Prewitt proper function by the request from Myrsloik.
Since I'm not familiar with SIMD, I may require time for beginning it.

[Chikuzen]

Updated to 0.2.0

* Add new function 'Sobel'.
* Add new function 'Prewitt'.
* Convolution/HV: add new option 'saturate'.
* Minimum/Maximum: add new option 'coordinates'
* various bug fix.

[Revgen]

The coordinates option and saturate option fix the issues HolyWu described. Thank You.

[Chikuzen]

But the output of generic.Sobel/generic.Prewitt is quite different from the equivalent mode of mt_edge.

from masktools2 docs
"sobel" uses the kernel "0 -1 0 -1 0 1 0 1 0"
I've never heard calling this "Sobel" except masktools.

"hprewitt" is equivalent to : mt_logic(mt_edge("1 2 1 0 0 0 -1 -2 -1 1"), mt_edge("1 0 -1 2 0 -2 1 0 -1 1"), mode = "max"),
hprewitt? I think this is Sobel (http://en.wikipedia.org/wiki/Sobel_operator)
and, Prewitt that I wrote is 'Prewitt compass edge detection' which requires 8 kernels.

[Myrsloik]

from masktools2 docs

I've never heard calling this "Sobel" except masktools.


hprewitt? I think this is Sobel (http://en.wikipedia.org/wiki/Sobel_operator)
and, Prewitt that I wrote is 'Prewitt compass edge detection' which requires 8 kernels.

Now you made me go check the reference literature...

Sobel is indeed defined as 1 2 1, 0 0 0, - 1 -2 - 1 and then the horizontal and vertical filters are usually summed up as sqrt(sh^2+sv^2) or abs(sh) + abs(sv) for speed. There are also variants where the filter is rotated when looking for diagonal edges but I've never seen it used.

Now on to prewitt. Apparently the prewitt guy was busy and suggested several similar things. But the one you want is described here: http://en.wikipedia.org/wiki/Prewitt_operator
and should be just about identical to sobel except that it's 1 1 1.

I really can't find any mention of masktool's so called sobel filter. I can however tell you this, it really is implemented like a single 0 1 0, 1 0 -1, 0 -1 0 convolution and I think it looks like a horrible idea. It's easy to see that edges get a VEEEEEERY different weight depending on directionality. Just imagine a diagonal gradient in the \ and / directions and see the difference in result.

[Chikuzen]

I found some typo and wrong understanding on Sobel and Prewitt.

updated to 0.2.1

* Sobel/Prewitt: fix invalid logic
* Sobel/Prewitt: Change default values of 'min' and 'max'

[aegisofrime]

So is there an updated version of QTGMC implementing these generic filters yet?

[Revgen]

^Not yet.

The current Vapoursynth QTGMC script is here. http://forum.doom9.org/showthread.php?t=166582

[Chikuzen]

updated to 0.2.2

* Sobel, Prewitt: add new option 'rshift'
* Sobel, Prewitt: fix 9/10/16bit format processing

[Are_]

Link to 0.2.2 does not work.

[Chikuzen]

Link to 0.2.2 does not work.

fixed

[Chikuzen]

updated to 0.3.0

* add new function 'Blur'

[kolak]

:thanks:

[kolak]

Looks like there is a bug in blur when format is 16bit- footage has some green artefacts (10bit and 8bit are ok).

[Chikuzen]

(Probably,) fixed.
genericfilters_0.4.1.dll (http://www.mediafire.com/download.php?m4w138a7m3xdxyy)

[kolak]

Yes- seams to be fine. Thanks.

Could someone port yadifmod (eedi3+nnedi3 are already there)? QTGMC won't come any soon, so at least we could have this for now.
AVS dll works, but only in yv12 mode and overall processing is 2x slower then in avs.

[Chikuzen]

update to 0.4.2

@HolyWu
probabry fixed.

[mastrboy]

What is the substitute to mt_merge?

[Myrsloik]

What is the substitute to mt_merge?

MaskedMerge() which is a part of the core.

[mastrboy]

MaskedMerge() which is a part of the core.

Ah, no wonder i didn't see it in the function list in the first post in this thread... Thanks...

[Chikuzen]

update to 0.5.0

* add new function 'Hysteresis'
* Convolution: increase maximum matrix length from 5 to 17 when mode is 'h' or 'v'
* various bug fix.


OFF TOPIC:
BTW, I referd to Masktools-v1's code first and wrote an avisynth2.6 plugin first, in order to understand the algorithm of hysteresis mask.
http://pastebin.com/37gPmkNy
It seems that this filter is faster than masktools2a48 about 150% to 300% on my machine :eek:
(Probably, since the lump of superfluous flesh called STL was eliminated from the critical part.)

[jackoneill]

@Chikuzen: What do you think about adding the edge mask algorithms from tedgemask?

[Chikuzen]

@Chikuzen: What do you think about adding the edge mask algorithms from tedgemask?

I've never read the code of TEdgeMask.
if it's possivel , i'll write.

[cretindesalpes]

IIRC mt_edge("prewitt") does a pretty similar but slightly better job than TEdgeMask. This can be noticed on very faint lines. Maybe just a different rounding behaviour. I don't know which one is the fastest.

[mandarinka]

Tedgemask's default thresh is fairly high (conservative). You need to set threshy to like 4 (well, from the anime dvd I am trying now, it seems the equivalent is between 3 and 4) if you want sensitivity of "mt_edge("prewitt")", it seems to me. Ofc, with that sensitivity, both pick up a lot of false positive pixels.

Hard to say (and I am lazy to check this too much), but I like tedgemask more I think.

[Chikuzen]

I wrote TEMmod to understand how TEdgeMask works.

http://forum.doom9.org/showthread.php?t=168390

Now, I'm going to drop type1 and type2 when I port this to VS since it seems that there is no difference not much.

Any suggestions?

[Chikuzen]

update to 0.6.0

* add new function 'TEdge'

[mastrboy]

update to 0.6.0

* add new function 'TEdge'

Thanks...
That would be the same as this for vapoursynth: http://forum.doom9.org/showthread.php?t=168390 ?

Any chance you're going to add tcanny mask as well?
Would be great to see most of the masking tools/algorithms collected into 1 plugin.

[sl1pkn07]

fail build. missing convolution_utils.h

greetings

[Chikuzen]

fail build. missing convolution_utils.h

greetings

fixed

[sl1pkn07]

thanks

[Reel.Deel]

Out of curiosity (and learning purposes) how would one implement the "undot6" plugin talked about here (http://mechaweaponsvidya.wordpress.com/2013/01/)?

[Chikuzen]

update to 0.7.0

* add new funtion 'GBlur'
* add new function 'Canny'

[Chikuzen]

Hi Chikuzen,

May I know do you plan to add 'mode' to generic.Canny like AviSynth version has(people use mostly mode=1 IMHO)? :thanks:

no.
I don't know the reason why people use Canny to get an edge mask instead of Sobel.

[Myrsloik]

Forked source with fixes is available from:
https://github.com/myrsloik/GenericFilters

Fixed windows binaries are bundled with the R22 (and later) VapourSynth installer.

[Mystery Keeper]

VirtualDub crashed on Canny.
canny = core.generic.Canny(z)
Crash report. (http://paste.org.ru/?75mhif)

[Myrsloik]

VirtualDub crashed on Canny.
canny = core.generic.Canny(z)
Crash report. (http://paste.org.ru/?75mhif)

Was that with the version included in the VS installer?
What was the format and bitdepth of the input to canny?

[Mystery Keeper]

It was. Format is common YV12. Right after decoding d2v project.

[Mystery Keeper]

Canny gives me 3 masks, one for each plane. How can I mix them all in one plane?

[Myrsloik]

Canny gives me 3 masks, one for each plane. How can I mix them all in one plane?

Either use two Lut2 to get the max/sum of the masks (depending on preference) or simply only use the luma mask for everything. Personally I'd only use the luma mask and be done with it...

[Mystery Keeper]

Tried looking at separate planes. Guess you're right.

[mawen1250]

There's 1 pixel up-shift in each plane while comparing the result of generic.Canny(chroma=True) with that of tcanny(plane=7)/TCannyMod(chroma=1).

[Myrsloik]

There's 1 pixel up-shift in each plane while comparing the result of generic.Canny(chroma=True) with that of tcanny(plane=7)/TCannyMod(chroma=True).

I'll take a look at it when I have time again. Are you sure it isn't tcanny that shifts it?

[mawen1250]

I made a simple test pattern and confirm that the result of generic.Canny is wrong.

[Myrsloik]

I made a simple test pattern and confirm that the result of generic.Canny is wrong.

Here's a possibly fixed 32 bit dll (https://www.dropbox.com/s/wd6cum67qrtb70q/GenericFilters.dll?dl=1). Not tested but probably working.

Should fix the canny line shift and several problems in the minimum/maximum filters as well. Report your findings. Don't forget to delete the dll that comes with VapourSynth.