Граф коммитов

  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • ea774e6 (HEAD -> master, trunk) build: a few Win32 build fixes. by sam 2014-06-13 07:23:18 +0000
  • e1ec73c build: update build system. by sam 2014-06-13 05:41:12 +0000
  • c117c13 autocontrast: fix alpha handling. by sam 2014-06-13 05:41:07 +0000
  • 027ac83 Allow to choose between RGB and YUV in pipi_open_sequence(). by sam 2011-02-08 22:09:53 +0000
  • d047d4b Reactivate pipi_dither_24to16, it's wrong but useful. by sam 2011-02-01 20:04:25 +0000
  • 0a6c2d8 Only use first stream in the storyboard code. by sam 2011-02-01 19:47:55 +0000
  • af83f81 YUV support in dumpmovie and makemovie. by sam 2010-11-03 21:31:38 +0000
  • 72505ff Ensure dumpmovie outputs YUVA pictures. by sam 2010-11-03 21:31:31 +0000
  • 01fc592 Import x264 presets into sequence.c. Doesn't work properly yet. by sam 2010-10-19 16:33:16 +0000
  • e88e7d9 Clamp Y to 16 - 235, as the MPEG standard uses. by sam 2010-10-19 16:33:10 +0000
  • 449823b Add a simple dumpmovie program to extract YUV images from a movie. by sam 2010-10-19 16:33:04 +0000
  • bd77b09 Add functions to convert from RGB to YUV and back. by sam 2010-10-17 14:33:57 +0000
  • 3f31901 Implement bicubic resampling. Lacks some blurring in the pre-pass, maybe. by sam 2010-10-17 00:57:27 +0000
  • 4ef27f0 Minor compilation warning fix. by sam 2010-10-16 21:55:44 +0000
  • 2297613 Make the movie generation methods and program more versatile. by sam 2010-10-16 21:53:01 +0000
  • 926f50d Update opencv codec to more recent cvSaveImage prototype. by sam 2010-10-16 21:52:56 +0000
  • 8c22661 Look for highgui.h in addition to the opencv package. by sam 2010-10-16 21:52:51 +0000
  • 231d51f Avoid floating-point divisions in the Bresenham resize. by sam 2010-10-16 21:52:45 +0000
  • b72179a Move resize.c to an algorithm-specific file so that we can implement other resizing methods. by sam 2010-10-16 21:52:38 +0000
  • adb764b On Win32, define 32-bit C99 types as ints instead of longs, because a lot of software (wrongly) assumes that int32_t is the same as int. by sam 2010-04-22 13:55:41 +0000
  • 9967e7c * Fixed Mac OSX compilation and Xcode project file by jylam 2009-12-01 08:54:07 +0000
  • f7d85c9 Add missing exec_prefix entry in pipi.pc.in. by sam 2009-08-25 14:48:34 +0000
  • 091a50a Update bootstrap script (automake 1.11 and pkg-config detection). by sam 2009-08-25 14:23:46 +0000
  • 2b8697c Fix a 32-bit compilation issue in img2twit. by sam 2009-08-25 13:49:44 +0000
  • 6d12c28 Mark pipi_open_sequence's first argument as const. by sam 2009-08-05 14:37:02 +0000
  • 5672006 Update the Linux build, following the makemovie code merge. by sam 2009-08-05 14:10:02 +0000
  • 4f228c2 makemovie no longer depends on FFmpeg libraries. by sam 2009-08-05 14:05:12 +0000
  • 23901eb Move the whole makemovie logic to libpipi. It's really hackish for now, but it seems to work. by sam 2009-08-05 14:03:37 +0000
  • ba926da Add USE_FFMPEG to the list of configure macros. by sam 2009-08-05 13:44:37 +0000
  • 2940aeb Do not bail out if only GDI+ is found. by sam 2009-08-03 23:21:36 +0000
  • a7486b1 Allow makemovie to build on Win32 using our ffmpeg binaries. by sam 2009-08-03 16:11:22 +0000
  • af1992d Add Win32 ffmpeg binaries for easier builds. by sam 2009-08-03 15:06:52 +0000
  • 5fb0f89 Do not touch the alpha layer in pipi_contrast(). by sam 2009-07-31 08:24:58 +0000
  • d004ff0 Fix the FFmpeg libraries detection. by sam 2009-07-29 21:34:12 +0000
  • 0448bc8 Fix build without CGAL. by sam 2009-07-29 09:18:27 +0000
  • 5d21969 That optimisation sucked. Reverted median filter to something that works both on Linux and Windows. by sam 2009-07-22 13:39:07 +0000
  • 3f6424f Improve the median filter performance. by sam 2009-07-22 12:09:34 +0000
  • db1fd94 Update the Win32 solution for Visual Studio 2008. by sam 2009-07-22 08:24:52 +0000
  • 48e7d48 Add a .desktop file for The Pimp. by sam 2009-07-22 07:14:11 +0000
  • 699614f Add missing blit.c to repository. by sam 2009-07-22 07:14:03 +0000
  • c89eae3 Don't force MJPEG in makemovie. Other codecs work just fine. by sam 2009-07-08 23:38:52 +0000
  • fb7d190 Implement pipi_blit(). by sam 2009-07-08 23:38:46 +0000
  • 859e659 Implement pipi_rotate(). Currently nearest-neighbour only, supersample if you need more precision. by sam 2009-07-08 23:38:39 +0000
  • f0b0a8b Fix the pipi_merge() semantics so that it matches Gimp’s, and don’t ignore the image’s alpha layers. by sam 2009-07-08 23:38:32 +0000
  • b473f55 Do not touch alpha values when modifying brightness or contrast. by sam 2009-07-08 23:38:27 +0000
  • 7cf0980 Look for uninstalled ffmpeg binaries before the system-wide ones. by sam 2009-07-08 23:38:17 +0000
  • d6e4f6d img2twit: reinstate old interpolation tuning for final picture quality. by sam 2009-05-30 16:01:29 +0000
  • 51036f8 img2twit: remove contrast enhancement. by sam 2009-05-30 16:01:23 +0000
  • e26e859 Make img2twit's points-per-cell value changeable at runtime. by sam 2009-05-30 16:01:17 +0000
  • 7a49a33 img2twit: prevent two points from having the same coordinates, since this is forbidden by the image format; simplify operation computations by not packing coordinates at runtime; add a version information in the file format for future extension. by sam 2009-05-30 16:01:09 +0000
  • 470a28b Optimise img2twit's cell packing even more, gaining around 60 bits. by sam 2009-05-28 01:54:56 +0000
  • a81278a Beginning of the optim... by sam 2009-05-28 01:54:50 +0000
  • 03f16f2 Add -Wextra to the list of compilation warnings. We're no longer warning-free, but there's not much to fix. by sam 2009-05-28 01:54:43 +0000
  • 14c3888 img2twit: change the position of the S component to help decompose points for a forthcoming optimisation. by sam 2009-05-28 01:54:35 +0000
  • fa9efdf Another img2twit space optimisation: if we have remaining bits at the end of the cell allocation, we try to use them to improve the RGB precision as well as the inner-grid coordinates. by sam 2009-05-27 05:04:48 +0000
  • b9e07e7 img2twit: store w+1,h+1 instead of w,h since neither width nor height can be zero. by sam 2009-05-27 05:04:41 +0000
  • 8c39330 Minor cleanup in img2twit. by sam 2009-05-27 05:04:35 +0000
  • ff93399 Fix img2twit's rendering. Apparently CGAL's Delaunay triangulation module does not work with negative indices. by sam 2009-05-27 05:04:30 +0000
  • 8ff3640 Increase img2twit's output contrast when rendering the final image. by sam 2009-05-27 05:04:23 +0000
  • efac3cf Autodetect message size in img2twit. by sam 2009-05-27 05:04:17 +0000
  • 858b677 Add multiple charsets support to img2twit, and autodetect charset when decoding. by sam 2009-05-27 05:04:10 +0000
  • 014e889 Allow for non-integer quality values in img2twit. by sam 2009-05-25 07:05:13 +0000
  • d18e224 Fix a bug in img2twit's cell allocation. by sam 2009-05-25 06:50:53 +0000
  • 901c4cd Make img2twit message length configurable at runtime, improve the initial guess by smoothing hard colour gradients, and fixed the bitstack copy constructor to avoid corruption on exit. by sam 2009-05-25 00:17:02 +0000
  • 6a02de1 Make img2twit more configurable: iterations per second and debug mode can now be chosen at runtime. Message length will soon be. by sam 2009-05-25 00:16:56 +0000
  • 9d2ab87 Add proper Unicode handling to img2twit. This includes an UTF-8 encoder/decoder and a special bit stack that can handle arbitrary bases within its stream. The image encoder and decoder are now separate code paths. by sam 2009-05-25 00:16:50 +0000
  • 1779efb Tremendously improve img2twit.cpp: only compute the new error around the updated pixel, compute bit allocation at runtime, weight operations in order to try usual "winners" more often, increase RGB allocation size. by sam 2009-05-25 00:16:44 +0000
  • 8bf0131 Add C++ warnings to the build process. by sam 2009-05-25 00:16:38 +0000
  • 4dd30b4 First attempt at a super-compressor for the purpose of sending images to Twitter, rendering this service slightly more useful. It's still full of crap, of course. by sam 2009-05-25 00:16:32 +0000
  • dbd1afd pipi_measure_msd(): fix this function so that it works with colour images, too. by sam 2009-05-25 00:16:26 +0000
  • d1fa1af Fix the FFmpeg headers location in example programs. by sam 2009-05-25 00:16:20 +0000
  • 335ba42 Look for the awesome CGAL library at configure stage. by sam 2009-05-25 00:16:13 +0000
  • d641902 Update ignore properties. by sam 2009-04-21 09:30:03 +0000
  • c7ad1f6 Synchronise Visual Studio projects with the current code. by sam 2009-04-21 09:29:03 +0000
  • e7c1b11 Add missing M_PI definition in wave.c for Win32 compilation. by sam 2009-04-20 23:57:14 +0000
  • 4c9c6fb Make sure that the gdiplus codec initialises GdiPlus before any library calls are made. by sam 2009-04-20 23:57:08 +0000
  • 25eac66 wave.c: implement pipi_wave() in addition to pipi_sine(). The effect is not yet satisfying. by sam 2009-04-20 23:57:02 +0000
  • 44eca2e context.c: factor some code to gain 180 lines of code. by sam 2009-04-20 23:56:56 +0000
  • 78c5fad wave.c: rename pipi_wave() into pipi_sine(). by sam 2009-04-20 23:56:50 +0000
  • 3b38443 resize.c: add pipi_crop() and the --crop flag. by sam 2009-04-20 23:56:41 +0000
  • f025663 jpeg.c: fix a huge resource leak in the error codepath of the JPEG codec. by sam 2009-03-28 23:26:47 +0000
  • 9ebe671 makemovie: new example program; this utility reads a list of images from stdin and encodes them into a movie (name specified on the command line). by sam 2009-03-28 23:26:38 +0000
  • 80bfef8 mean.c: rename this file into merge.c and implement pipi_merge() for trivial image merging. by sam 2009-03-24 00:49:00 +0000
  • b7e3368 wave.c: change the wave arguments. The format is now eg. 150x100+10r1.3 where 150 is the wave period, 100 is the amplitude, 10 is the phase, and 1.3 is the angle of the wave direction. by sam 2009-03-24 00:48:53 +0000
  • 9e9d93d histogram.c: fix a shadowing declaration. by sam 2009-03-24 00:48:44 +0000
  • 37a1e56 wave.c: better behaviour around image edges. by sam 2009-02-21 14:26:58 +0000
  • 4936ebf Fix a very old crash in the box blur filter that occurred when the image was broader than tall. by sam 2009-02-21 14:26:51 +0000
  • 05c7be5 Fix a crash in the JPEG loader when meeting with ENOENT. by sam 2009-02-21 14:26:45 +0000
  • 7811f18 storyboard.c: fix compilation with optimisations disabled. by sam 2009-02-21 14:26:38 +0000
  • f0b30c6 Properly handle alpha components in the convolution code. by sam 2009-02-21 14:26:32 +0000
  • 1b42147 Properly handle alpha components in the resize code. by sam 2009-02-21 14:26:24 +0000
  • 2ff222a Add a wave filter for crappy sine effects. by sam 2009-02-21 14:26:01 +0000
  • e0e1a1f storyboard.c: add decorations around thumbnails by sam 2009-01-16 14:23:23 +0000
  • 48e2c4c storyboard.c: - generate smaller thumbnail mosaics - process more images (1 out of 15 instead of 20) - try to rule out more similar images - cope with streams that do not advertise their picture size in the headers by sam 2009-01-16 14:23:14 +0000
  • 78cc659 imlib.c: fix support for alpha channel when saving using Imlib2. by sam 2009-01-15 21:36:35 +0000
  • d759c65 storyboard.c: more output filename tuning... by sam 2009-01-13 10:38:26 +0000
  • 0e8b0d6 storyboard.c: fix output path. by sam 2009-01-13 10:34:17 +0000
  • 7701640 storyboard.c: tuned the output file name. examples/.gitignore: ignore storyboard. by sam 2009-01-13 10:29:29 +0000
  • f83b0de Conditionally build the movie thumbnailer by detecting libavcodec et al. by sam 2009-01-12 23:52:00 +0000
  • 70e7f92 Added a movie thumbnailer (storyboard generator) using libavcodec. by sam 2009-01-12 23:51:52 +0000