Search found 39 matches

by gubach
2014-12-21T08:23:02-07:00
Forum: PerlMagick
Topic: extracting a shape from a transparent background efficiently
Replies: 3
Views: 7384

Re: extracting a shape from a transparent background efficiently

Thaks for the answers. I only want to remove all extra transparent background after rotation. Version is: ImageMagick 6.7.8-7 2013-07-10 Q8 I looked in the description of the (parameterless) method: Trim remove edges that are the background color from the image and this means a usage like $image=>Tr...
by gubach
2014-12-20T04:21:18-07:00
Forum: PerlMagick
Topic: extracting a shape from a transparent background efficiently
Replies: 3
Views: 7384

extracting a shape from a transparent background efficiently

Suppose an image is given consisting of a monochrome shape (polygon) and a transparent background; there is no transparent border around the shape; points or edges touching the image border. If such images are rotated the ratio of the image may increase and the shape is not any more touching the ima...
by gubach
2014-07-06T09:11:08-07:00
Forum: PerlMagick
Topic: deleting duplicates in an image list
Replies: 1
Views: 5202

deleting duplicates in an image list

I have a list with downscaled black-and-white images ($mask_segment_list_dual) generated by an image processing pipeline and I want to delete duplicates in the list. In the command line tools I found compare -metric AE (or MAE) image1.png image2.png null: 2>&1 that delivers 0 if the two images are t...
by gubach
2014-02-25T07:52:46-07:00
Forum: Users
Topic: Edge() with unacceptable artefacts
Replies: 8
Views: 6812

Re: Edge() with unacceptable artefacts

fmw42 wrote: $image->Morphology(kernel=>'edgein diamond:1', channel=>"all");
no, I got a "Failed to parse kernel number #0" error and image is unchanged.
by gubach
2014-02-24T11:08:29-07:00
Forum: Users
Topic: Edge() with unacceptable artefacts
Replies: 8
Views: 6812

Re: Edge() with unacceptable artefacts

I tried a workaround where I write the image, construct a $systemstring with "-morphology edgein diamond:", call the system from Perl with $systemstring and Read the result image to work with it in my processing pipeline. After some initial tries this ugly workaround is now working, but I posted a q...
by gubach
2014-02-24T08:45:02-07:00
Forum: PerlMagick
Topic: how to map "-morphology edgein diamond:1" in PerlMagick
Replies: 0
Views: 4933

how to map "-morphology edgein diamond:1" in PerlMagick

Hello, I want to map this command line convert imput.png -morphology edgein diamond:1 result.png in PerlMagick but with the description of the Morphology method Morphology kernel=>string, channel=>{All, Default, Alpha, Black, Blue, CMYK, Cyan, Gray, Green, Index, Magenta, Opacity, Red, RGB, Yellow},...
by gubach
2014-02-23T05:17:13-07:00
Forum: Users
Topic: Edge() with unacceptable artefacts
Replies: 8
Views: 6812

Re: Edge() with unacceptable artefacts

convert 12701896523_c48135e0f9_k.jpg -morphology edgein diamond:1 result.png yes this works much!! better. Thank you very much! Now I need to integrate it in PerlMagick, but the description of the method Morphology on http://www.imagemagick.org/script/perl-magick.php Morphology kernel=>string, chan...
by gubach
2014-02-22T12:12:02-07:00
Forum: Users
Topic: Edge() with unacceptable artefacts
Replies: 8
Views: 6812

Edge() with unacceptable artefacts

Hello, I want to make edge images from geometric gray scale images (with a Perlmagick subprogram) but I get unacceptable artefacts. I used the edge-function in non geometric contexts a lot and I did not care about the artefacts there but in the geometric context it becomes unacceptable. Is the edge-...
by gubach
2013-12-13T11:56:39-07:00
Forum: PerlMagick
Topic: p6m tile perlmagick subprogram problem
Replies: 0
Views: 4404

p6m tile perlmagick subprogram problem

Hello, some years ago I wrote perl subprograms for the 17 eucledian symmetry groups. They worked sufficient for my art context but the groups based on non 90° and 45° angles have some minor artefacts. For one typ of those artefacts (black pixels at the edges were two triangles were composed) I have ...