223 Commits (3bf0d166a84377d50ba08ad20eefc8bde9539e27)
 

Author SHA1 Message Date
  sam 77ec441434 * blur.c: implement box blur; currently runs in O(n) but we could make it 16 years ago
  sam b7c2d1f847 * autocontrast.c: simple autocontrast filter; does not work very well. 16 years ago
  sam c269c219b2 * Get rid of test.c, it was no longer useful anyway. 16 years ago
  jylam baa61a9f6d * Handle alpha layer in floodfill (but don't make it conditionnal to the algorithm) 16 years ago
  jylam 290e151913 * First shot of a floodfiller (both u32 and float, 4 neighbours) 16 years ago
  sam 53268650a7 * dbs.c: generate the initial halftone using random dithering instead of 16 years ago
  sam 0eda956bfc * random.c: implement random dithering using a deterministic pseudo-RNG. 16 years ago
  sam 18dd51e1a5 * Dithering algorithms no longer modify the original image. 16 years ago
  sam 2dc8bb35dd * pipi.c: fix a memory leak caused by empty picture having codec_format 16 years ago
  sam 66f9ce5ea4 * pipi.c: implement pipi_copy(). 16 years ago
  sam a17c6753e3 * pixels.c: store byte length and bits-per-pixel value in the pipi_pixels_t 16 years ago
  sam 41efcdecc0 * Prefix dithering functions with _dither_ to avoid namespace cluttering. 16 years ago
  sam 6a3309663e * ordered.c: implement Bayer dithering (pretty trivial). 16 years ago
  sam a6dcd80490 * dbs.c: optimise DBS by ignoring 16x16 cells that had no pixel changes for 16 years ago
  sam 2a3c7ca990 * dbs.c: improve the DBS human visual system kernel by adding two Gaussian 16 years ago
  sam bf31390d1e * sharpen.c: add a sharpen filter example, using our generic convolution 16 years ago
  sam 8c4591f1d9 * blur.c: remove the blurring code and use our generic convolution 16 years ago
  sam 03ca8dc76d * convolution.c: automatically detect when a convolution filter is separable 16 years ago
  sam c587c29021 * blur.c: fix the blur example’s argument checking. 16 years ago
  sam 89f35cb8fd * convolution.c: generic convolution method. Does not take advantage of 16 years ago
  sam 870828bc87 * edd.c: output MSD instead of RMSD in the displacement computation. 16 years ago
  sam 87f8e922ab * Error diffusion methods now support either raster or serpentine scan. 16 years ago
  sam 99d8af784e * pixels.c: support more conversion combinations. 16 years ago
  sam 302ce37f8e * ostromoukhov.c: Ostromoukhov's simple error diffusion algorithm. 16 years ago
  sam 2182fdfdbf * floydsteinberg.c: perform Floyd-Steinberg dithering on a serpentine path. 16 years ago
  sam 3e575ba28a * dither.c: add an example program for dithering methods. 16 years ago
  sam 628c2618da * dbs.c: new dithering function: Direct Binary Search. One of the best 16 years ago
  sam eaad4db6f9 * measure.c: there is now pipi_measure_msd in addition to pipi_measure_rmsd. 16 years ago
  sam 18fd8276bd * edd.c: output E_fast as well. 16 years ago
  sam 5974ab95c5 * edd.c: example program that computes the Floyd-Steinberg displacement, 16 years ago
  sam 4a278923ff * blur.c: adapt the kernel size to large values of dx and/or dy. 16 years ago
  sam 1408cff41d * Add Floyd-Steinberg grayscale dithering. 16 years ago
  sam 36da8b05dd * pixels.c: fix a typo in the dithering method that could cause crashes. 16 years ago
  sam 9a536afd4f * measure.c: started writing error/measure functions. First one is RMSD. 16 years ago
  sam a11f92ff95 * blur.c: support for greyscale images. 16 years ago
  sam eea9f9b829 * pixels.c: start supporting grayscale images. 16 years ago
  sam f687640a09 * codec.c: bump Imlib2 and OpenCV priorities over SDL. 16 years ago
  sam af15db52cb * opencv.c: bring the OpenCV codec up to date. 16 years ago
  sam 0d8f01566b * pixels.c: add support for 24-bpp BGR format. 16 years ago
  sam 8953c91d6b * pipi.c: reimplement pipi_new() without relying on the underlying codec 16 years ago
  sam 732f6b2335 * configure.ac: fix the OpenCV detection by using pkg-config. 16 years ago
  sam 3d86c2f45a * Start shipping unit tests. The first one checks that converting an 16 years ago
  sam caba836e5d * pixels.c: clamp pixels in the float32 -> rgba32 conversion to avoid 16 years ago
  sam 877e7b21a9 * filter/blur.c: avoid annoying side effects with very small or negative 16 years ago
  sam ea38654c88 * filter/blur.c: minor optimisation in the kernel generation. 16 years ago
  sam 4a857277a7 * blur.c: support for variable blur radius and Gaussian offset. 16 years ago
  sam e15f20a316 * pixels.c: add Floyd-Steinberg dithering to the float32 -> rgba32 16 years ago
  sam 0105b6f3b6 * pixels.c: speed up the RGBA32 -> float32 conversion, using a simple LUT. 16 years ago
  sam 42af43ad99 * blur.c: separated blur; it's now incredibly faster. 16 years ago
  sam 19874f4650 * Free surfaces allocated by pipi_getpixels() in pipi_free(). 16 years ago