 Dirk Lemstra
|
14d259beff
|
Minor optimization.
|
2025-04-10 17:17:41 +02:00 |
|
 Cristy
|
21e324b9d9
|
respect image bounds
|
2025-04-07 19:27:32 -04:00 |
|
|
Cristy
|
2e59567657
|
you must explicitly set compare:virtual-pixel=false
|
2025-04-06 21:35:02 -04:00 |
|
|
Cristy
|
7154d1ecb1
|
ebale virtual pixels by default
|
2025-04-06 21:18:08 -04:00 |
|
|
Cristy
|
07d1ba312e
|
support -define compare:virtual-pixels=true define
|
2025-04-06 19:20:00 -04:00 |
|
|
Cristy
|
741b0ffd55
|
don't close a dash array subpath
|
2025-04-06 10:43:09 -04:00 |
|
|
Cristy
|
a4c963a7e2
|
cosmetic
|
2025-04-06 09:20:03 -04:00 |
|
|
Dirk Lemstra
|
84c2b58531
|
Removed variable that was only assigned but not used.
|
2025-04-05 15:04:12 +02:00 |
|
|
Dirk Lemstra
|
07a9f6c493
|
Added CastDoubleToUShort and used it in the coders.
|
2025-04-02 17:52:51 +02:00 |
|
|
Dirk Lemstra
|
6f23697026
|
Added CastDoubleToUInt and used it in the coders.
|
2025-04-02 17:44:20 +02:00 |
|
|
Dirk Lemstra
|
b561299c7f
|
Use CastDoubleToSizeT to make sure we avoid invalid values when casting from a double.
|
2025-04-02 17:43:59 +02:00 |
|
|
Dirk Lemstra
|
cbe4770f0c
|
Renamed CastDoubleToLong to CastDoubleToSsizeT.
|
2025-04-01 22:45:11 +02:00 |
|
|
Dirk Lemstra
|
fd418ddc44
|
.Renamed CastDoubleToUnsigned to CastDoubleToSizeT
|
2025-04-01 22:34:12 +02:00 |
|
 Frank Dana
|
14a434f278
|
magick-config.h: Remove redundant block (#8076)
|
2025-03-31 23:16:13 +02:00 |
|
|
Cristy
|
c8f4e8cb75
|
render closed sub-path poly lines
|
2025-03-29 13:12:08 -04:00 |
|
|
Cristy
|
981ae2db2a
|
correct PSNR metric
|
2025-03-29 10:54:55 -04:00 |
|
|
Cristy
|
ddf1f5576b
|
correct PSNR distortion
|
2025-03-25 20:11:14 -04:00 |
|
|
Cristy
|
e94404b962
|
revert
|
2025-03-23 16:08:59 -04:00 |
|
|
Cristy
|
432c63e773
|
initialize channel minimum
|
2025-03-23 15:54:52 -04:00 |
|
|
Cristy
|
d1208c1cc3
|
negate AE/PAE/MEPP correlation image
|
2025-03-22 16:18:21 -04:00 |
|
|
Cristy
|
45c90d2c1a
|
reverse PSNR edge cases
|
2025-03-21 22:24:07 -04:00 |
|
|
Cristy
|
6687c754da
|
set PSNR for identical images to 48.1647db
|
2025-03-21 19:10:21 -04:00 |
|
|
Cristy
|
4678062d9b
|
PSNR metric returned incorrect results for spatial, FFT worked fine
|
2025-03-16 17:33:46 -04:00 |
|
|
Cristy
|
e6be2579b4
|
revert
|
2025-03-16 12:21:03 -04:00 |
|
|
Cristy
|
1b0f75ec9f
|
respect filename:literal define for both input and output filenames
|
2025-03-16 12:10:47 -04:00 |
|
|
Cristy
|
934d3f949d
|
strip thumbnail URI of decorators
|
2025-03-16 11:49:49 -04:00 |
|
|
Cristy
|
f060e09087
|
special case where target and reconstruction are same size
|
2025-03-16 09:07:55 -04:00 |
|
|
Cristy
|
9224c0dc46
|
revert special case where target and reconstruction is the same dimensions
|
2025-03-16 08:52:27 -04:00 |
|
|
Cristy
|
8f23ea73e0
|
don't encode URI '/'
|
2025-03-15 19:53:51 -04:00 |
|
|
Cristy
|
fae1740b6c
|
restore output of PSNR metric
|
2025-03-15 14:04:08 -04:00 |
|
|
Dirk Lemstra
|
fb883b589c
|
Only divide the value once when scaling quantum to an unsigned char. This is probably already optimized by the compiler but this change makes our intent more clear.
|
2025-03-15 14:36:46 +01:00 |
|
|
Cristy
|
41304894af
|
for NCC metric, identical is 1 not 0
|
2025-03-15 09:29:07 -04:00 |
|
|
Cristy
|
62cb29b80a
|
fix issue with similarity search of two images of same dimensions
|
2025-03-15 08:12:22 -04:00 |
|
|
Cristy
|
66985e847f
|
cosmetic
|
2025-03-09 15:23:31 -04:00 |
|
|
Cristy
|
225b65d7fb
|
negate SSIM similarity image
|
2025-03-09 13:37:33 -04:00 |
|
|
Cristy
|
3cbce56961
|
negate SSIM similarity metric
|
2025-03-08 21:48:53 -05:00 |
|
|
Cristy
|
8f8cb86b76
|
cosmetic
|
2025-03-08 20:42:44 -05:00 |
|
|
Cristy
|
dd44681218
|
negate PHASH similarity image
|
2025-03-08 07:44:43 -05:00 |
|
|
Cristy
|
f02769cfcd
|
negate MAE metric
|
2025-03-07 22:06:31 -05:00 |
|
|
Dirk Lemstra
|
e8b195dd10
|
Corrected memset.
|
2025-03-07 11:48:03 +01:00 |
|
|
Dirk Lemstra
|
be01f28ae2
|
Added our own implementation of parsing an iso860 date and time to fix the Windows build.
|
2025-03-07 11:28:03 +01:00 |
|
 Mamoru TASAKA
|
ae34a44993
|
Fix statistic.c GetImageRange initializer (#8010)
https://github.com/ImageMagick/ImageMagick/commit/099e1a2ef7bf91f26f5281a165a85443b285abd9
cleaned up row_minima row_maxima initialization.
But with the above commit these 2 values are once initialized with
MagickMinimumValue, MagickMaximumValue, later these are again
initialized with p[0]. The latter is apparently not needed.
Fixes https://github.com/rmagick/rmagick/issues/1679 .
|
2025-03-06 10:26:23 -05:00 |
|
|
Cristy
|
f70d8d3759
|
support epoch() and magicktime() FX functions contributed by @snibgo
|
2025-03-05 17:52:30 -05:00 |
|
|
Cristy
|
7021e98772
|
include PNG tRNS chunk
|
2025-03-01 09:21:17 -05:00 |
|
|
Cristy
|
41e87e5124
|
optimized Oklab colorspace transformation
|
2025-02-27 19:32:25 -05:00 |
|
|
Cristy
|
14511d8b36
|
mime:type is not automatically injected, you must request it
|
2025-02-26 19:13:47 -05:00 |
|
|
Cristy
|
9656f70c48
|
file URI is we have an absolute path
|
2025-02-25 20:03:54 -05:00 |
|
|
Dirk Lemstra
|
14dfb98946
|
Also print the requested sizes in the error message when width or height exceeds the limits.
|
2025-02-25 19:38:41 +01:00 |
|
|
Cristy
|
4d249e1391
|
search for null, not 0
|
2025-02-25 10:03:08 -05:00 |
|
|
Cristy
|
7b415479ef
|
squash abort
|
2025-02-25 07:49:15 -05:00 |
|