Legacy ImageMagick Discussions Archive

I'm using a Python library that creates pngs, but once in a blue moon, generates a set of pngs that are essentially just transparent. I have no idea why and it's a rare enough event not to want to bother investigating.

The images are intended for social media, and I'd like to be able to test for transparency before they get sent & if there's a problem, we can rerun the image generator & alert the admin (me).

I found this snippet of code on a (very) old post, and it works perfectly (on Linux 32bit IM 6.9.10), but I wondered if there is a neater/quicker/whatever equivalent for both 6.9 and 7.x ?

Code: Select all

convert image -matte -channel A -separate -format '%[fx:image.maxima]' info:

Not really that I know. But your syntax is out of date. It should be

convert image -alpha on -channel A -format "%[fx:maxima]\n" info:

opaque:

Code: Select all

convert logo: -alpha on -channel a -separate -format "%[fx:maxima]\n" info:
1

partially transparent:

Code: Select all

convert logo: -transparent white logot.png
convert logot.png -alpha on -channel a -separate -format "%[fx:maxima]\n" info:
1

fully transparent:

Code: Select all

convert logo: -alpha on -channel a -evaluate set 0 +channel logot2.png
convert logot2.png -alpha on -channel a -separate -format "%[fx:maxima]\n" info:
0

If you want to detect partial transparency, change maxima to mean.

If you only want to check if fully opaque, then you can use

Code: Select all

convert image -format "%[opaque]\" info:

Code: Select all

convert logo: -format "%[opaque]\n" info:
true

Code: Select all

convert logot.png -format "%[opaque]\n" info:
false

Code: Select all

convert logot2.png -format "%[opaque]\n" info:
false

See https://imagemagick.org/script/escape.php

Thanks very much indeed.
Yes, definitely would have been out of date as I harvested that from a 2008 post on this board.

Legacy ImageMagick Discussions Archive

Testing for image transparency

Testing for image transparency

Re: Testing for image transparency

Re: Testing for image transparency