FFmpeg
vf_elbg.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2013 Stefano Sabatini
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 /**
22  * @file
23  * video quantizer filter based on ELBG
24  */
25 
26 #include "libavcodec/elbg.h"
27 #include "libavutil/opt.h"
28 #include "libavutil/pixdesc.h"
29 #include "libavutil/random_seed.h"
30 
31 #include "avfilter.h"
32 #include "drawutils.h"
33 #include "formats.h"
34 #include "internal.h"
35 #include "video.h"
36 
37 typedef struct ELBGFilterContext {
38  const AVClass *class;
39  struct ELBGContext *ctx;
41  int64_t lfg_seed;
43  int *codeword;
46  int *codebook;
49  uint8_t rgba_map[4];
50  int use_alpha;
51  int pal8;
53 
54 #define OFFSET(x) offsetof(ELBGFilterContext, x)
55 #define FLAGS AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
56 
57 static const AVOption elbg_options[] = {
58  { "codebook_length", "set codebook length", OFFSET(codebook_length), AV_OPT_TYPE_INT, { .i64 = 256 }, 1, INT_MAX, FLAGS },
59  { "l", "set codebook length", OFFSET(codebook_length), AV_OPT_TYPE_INT, { .i64 = 256 }, 1, INT_MAX, FLAGS },
60  { "nb_steps", "set max number of steps used to compute the mapping", OFFSET(max_steps_nb), AV_OPT_TYPE_INT, { .i64 = 1 }, 1, INT_MAX, FLAGS },
61  { "n", "set max number of steps used to compute the mapping", OFFSET(max_steps_nb), AV_OPT_TYPE_INT, { .i64 = 1 }, 1, INT_MAX, FLAGS },
62  { "seed", "set the random seed", OFFSET(lfg_seed), AV_OPT_TYPE_INT64, {.i64 = -1}, -1, UINT32_MAX, FLAGS },
63  { "s", "set the random seed", OFFSET(lfg_seed), AV_OPT_TYPE_INT64, { .i64 = -1 }, -1, UINT32_MAX, FLAGS },
64  { "pal8", "set the pal8 output", OFFSET(pal8), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, FLAGS },
65  { "use_alpha", "use alpha channel for mapping", OFFSET(use_alpha), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
66  { NULL }
67 };
68 
70 
72 {
73  ELBGFilterContext *const elbg = ctx->priv;
74 
75  if (elbg->pal8 && elbg->codebook_length > 256) {
76  av_log(ctx, AV_LOG_ERROR, "pal8 output allows max 256 codebook length.\n");
77  return AVERROR(EINVAL);
78  }
79 
80  if (elbg->lfg_seed == -1)
81  elbg->lfg_seed = av_get_random_seed();
82 
83  av_lfg_init(&elbg->lfg, elbg->lfg_seed);
84  return 0;
85 }
86 
88 {
89  ELBGFilterContext *const elbg = ctx->priv;
90  int ret;
91 
92  static const enum AVPixelFormat pix_fmts[] = {
96  };
97  if (!elbg->pal8) {
99  } else {
100  static const enum AVPixelFormat pal8_fmt[] = {
103  };
104  if ((ret = ff_formats_ref(ff_make_format_list(pix_fmts), &ctx->inputs[0]->outcfg.formats)) < 0 ||
105  (ret = ff_formats_ref(ff_make_format_list(pal8_fmt), &ctx->outputs[0]->incfg.formats)) < 0)
106  return ret;
107  }
108  return 0;
109 }
110 
111 #define NB_COMPONENTS 4
112 
114 {
115  AVFilterContext *ctx = inlink->dst;
116  ELBGFilterContext *const elbg = ctx->priv;
117 
118  elbg->pix_desc = av_pix_fmt_desc_get(inlink->format);
119  elbg->codeword_length = inlink->w * inlink->h;
120  elbg->codeword = av_realloc_f(elbg->codeword, elbg->codeword_length,
121  NB_COMPONENTS * sizeof(*elbg->codeword));
122  if (!elbg->codeword)
123  return AVERROR(ENOMEM);
124 
127  sizeof(*elbg->codeword_closest_codebook_idxs));
129  return AVERROR(ENOMEM);
130 
131  elbg->codebook = av_realloc_f(elbg->codebook, elbg->codebook_length,
132  NB_COMPONENTS * sizeof(*elbg->codebook));
133  if (!elbg->codebook)
134  return AVERROR(ENOMEM);
135 
136  ff_fill_rgba_map(elbg->rgba_map, inlink->format);
137 
138  return 0;
139 }
140 
141 #define R 0
142 #define G 1
143 #define B 2
144 #define A 3
145 
147 {
148  ELBGFilterContext *const elbg = inlink->dst->priv;
149  int i, j, k, ret;
150  uint8_t *p, *p0;
151 
152  const uint8_t r_idx = elbg->rgba_map[R];
153  const uint8_t g_idx = elbg->rgba_map[G];
154  const uint8_t b_idx = elbg->rgba_map[B];
155  const uint8_t a_idx = elbg->rgba_map[A];
156 
157  /* build the codeword */
158  p0 = frame->data[0];
159  k = 0;
160  for (i = 0; i < inlink->h; i++) {
161  p = p0;
162  for (j = 0; j < inlink->w; j++) {
163  elbg->codeword[k++] = p[b_idx];
164  elbg->codeword[k++] = p[g_idx];
165  elbg->codeword[k++] = p[r_idx];
166  elbg->codeword[k++] = elbg->use_alpha ? p[a_idx] : 0xff;
167  p += elbg->pix_desc->nb_components;
168  }
169  p0 += frame->linesize[0];
170  }
171 
172  /* compute the codebook */
173  ret = avpriv_elbg_do(&elbg->ctx, elbg->codeword, NB_COMPONENTS,
174  elbg->codeword_length, elbg->codebook,
175  elbg->codebook_length, elbg->max_steps_nb,
176  elbg->codeword_closest_codebook_idxs, &elbg->lfg, 0);
177  if (ret < 0) {
179  return ret;
180  }
181 
182  if (elbg->pal8) {
183  AVFilterLink *outlink = inlink->dst->outputs[0];
184  AVFrame *out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
185  uint32_t *pal;
186 
187  if (!out) {
189  return AVERROR(ENOMEM);
190  }
193  pal = (uint32_t *)out->data[1];
194  p0 = (uint8_t *)out->data[0];
195 
196  for (i = 0; i < elbg->codebook_length; i++) {
197  const int al = elbg->use_alpha ? elbg->codebook[i*4+3] : 0xff;
198  pal[i] = al << 24 |
199  (elbg->codebook[i*4+2] << 16) |
200  (elbg->codebook[i*4+1] << 8) |
201  elbg->codebook[i*4 ];
202  }
203 
204  k = 0;
205  for (i = 0; i < inlink->h; i++) {
206  p = p0;
207  for (j = 0; j < inlink->w; j++, p++) {
208  p[0] = elbg->codeword_closest_codebook_idxs[k++];
209  }
210  p0 += out->linesize[0];
211  }
212 
213  return ff_filter_frame(outlink, out);
214  }
215 
216  /* fill the output with the codebook values */
217  p0 = frame->data[0];
218 
219  k = 0;
220  for (i = 0; i < inlink->h; i++) {
221  p = p0;
222  for (j = 0; j < inlink->w; j++) {
223  int cb_idx = NB_COMPONENTS * elbg->codeword_closest_codebook_idxs[k++];
224  p[b_idx] = elbg->codebook[cb_idx];
225  p[g_idx] = elbg->codebook[cb_idx+1];
226  p[r_idx] = elbg->codebook[cb_idx+2];
227  p[a_idx] = elbg->use_alpha ? elbg->codebook[cb_idx+3] : 0xFFu;
228  p += elbg->pix_desc->nb_components;
229  }
230  p0 += frame->linesize[0];
231  }
232 
233  return ff_filter_frame(inlink->dst->outputs[0], frame);
234 }
235 
237 {
238  ELBGFilterContext *const elbg = ctx->priv;
239 
240  avpriv_elbg_free(&elbg->ctx);
241 
242  av_freep(&elbg->codebook);
243  av_freep(&elbg->codeword);
245 }
246 
247 static const AVFilterPad elbg_inputs[] = {
248  {
249  .name = "default",
250  .type = AVMEDIA_TYPE_VIDEO,
252  .config_props = config_input,
253  .filter_frame = filter_frame,
254  },
255 };
256 
258  .name = "elbg",
259  .description = NULL_IF_CONFIG_SMALL("Apply posterize effect, using the ELBG algorithm."),
260  .priv_size = sizeof(ELBGFilterContext),
261  .priv_class = &elbg_class,
262  .init = init,
263  .uninit = uninit,
267 };
ff_get_video_buffer
AVFrame * ff_get_video_buffer(AVFilterLink *link, int w, int h)
Request a picture buffer with a specific set of permissions.
Definition: video.c:108
OFFSET
#define OFFSET(x)
Definition: vf_elbg.c:54
AVPixelFormat
AVPixelFormat
Pixel format.
Definition: pixfmt.h:64
ELBGFilterContext::codeword_closest_codebook_idxs
int * codeword_closest_codebook_idxs
Definition: vf_elbg.c:45
AVERROR
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later. That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another. Frame references ownership and permissions
opt.h
ff_make_format_list
AVFilterFormats * ff_make_format_list(const int *fmts)
Create a list of supported formats.
Definition: formats.c:401
out
FILE * out
Definition: movenc.c:54
av_lfg_init
av_cold void av_lfg_init(AVLFG *c, unsigned int seed)
Definition: lfg.c:32
u
#define u(width, name, range_min, range_max)
Definition: cbs_h2645.c:250
ff_filter_frame
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:978
av_pix_fmt_desc_get
const AVPixFmtDescriptor * av_pix_fmt_desc_get(enum AVPixelFormat pix_fmt)
Definition: pixdesc.c:2964
ELBGFilterContext::lfg_seed
int64_t lfg_seed
Definition: vf_elbg.c:41
ELBGFilterContext::lfg
AVLFG lfg
Definition: vf_elbg.c:40
inlink
The exact code depends on how similar the blocks are and how related they are to the and needs to apply these operations to the correct inlink or outlink if there are several Macros are available to factor that when no extra processing is inlink
Definition: filter_design.txt:212
A
#define A
Definition: vf_elbg.c:144
av_frame_free
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:100
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:340
pixdesc.h
AVOption
AVOption.
Definition: opt.h:251
FILTER_QUERY_FUNC
#define FILTER_QUERY_FUNC(func)
Definition: internal.h:169
AV_PIX_FMT_BGR24
@ AV_PIX_FMT_BGR24
packed RGB 8:8:8, 24bpp, BGRBGR...
Definition: pixfmt.h:69
AV_PIX_FMT_BGRA
@ AV_PIX_FMT_BGRA
packed BGRA 8:8:8:8, 32bpp, BGRABGRA...
Definition: pixfmt.h:95
G
#define G
Definition: vf_elbg.c:142
ELBGFilterContext::codeword_length
int codeword_length
Definition: vf_elbg.c:44
ELBGFilterContext::pix_desc
const AVPixFmtDescriptor * pix_desc
Definition: vf_elbg.c:48
AVFilter::name
const char * name
Filter name.
Definition: avfilter.h:170
video.h
AVFrame::data
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:361
ELBGFilterContext::rgba_map
uint8_t rgba_map[4]
Definition: vf_elbg.c:49
av_get_random_seed
uint32_t av_get_random_seed(void)
Get a seed to use in conjunction with random functions.
Definition: random_seed.c:167
formats.h
query_formats
static int query_formats(AVFilterContext *ctx)
Definition: vf_elbg.c:87
elbg_inputs
static const AVFilterPad elbg_inputs[]
Definition: vf_elbg.c:247
ELBGFilterContext::use_alpha
int use_alpha
Definition: vf_elbg.c:50
AVFilterPad
A filter pad used for either input or output.
Definition: internal.h:47
avpriv_elbg_do
int avpriv_elbg_do(ELBGContext **elbgp, int *points, int dim, int numpoints, int *codebook, int num_cb, int max_steps, int *closest_cb, AVLFG *rand_state, uintptr_t flags)
Implementation of the Enhanced LBG Algorithm Based on the paper "Neural Networks 14:1219-1237" that c...
Definition: elbg.c:462
AV_LOG_ERROR
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:180
av_cold
#define av_cold
Definition: attributes.h:90
ff_video_default_filterpad
const AVFilterPad ff_video_default_filterpad[1]
An AVFilterPad array whose only entry has name "default" and is of type AVMEDIA_TYPE_VIDEO.
Definition: video.c:36
ff_vf_elbg
const AVFilter ff_vf_elbg
Definition: vf_elbg.c:257
ELBGFilterContext::codebook
int * codebook
Definition: vf_elbg.c:46
ff_formats_ref
int ff_formats_ref(AVFilterFormats *f, AVFilterFormats **ref)
Add *ref as a new reference to formats.
Definition: formats.c:617
ff_set_common_formats_from_list
int ff_set_common_formats_from_list(AVFilterContext *ctx, const int *fmts)
Equivalent to ff_set_common_formats(ctx, ff_make_format_list(fmts))
Definition: formats.c:776
AV_OPT_TYPE_INT64
@ AV_OPT_TYPE_INT64
Definition: opt.h:226
pix_fmts
static enum AVPixelFormat pix_fmts[]
Definition: libkvazaar.c:304
ctx
AVFormatContext * ctx
Definition: movenc.c:48
FILTER_INPUTS
#define FILTER_INPUTS(array)
Definition: internal.h:192
AV_PIX_FMT_RGBA
@ AV_PIX_FMT_RGBA
packed RGBA 8:8:8:8, 32bpp, RGBARGBA...
Definition: pixfmt.h:93
frame
static AVFrame * frame
Definition: demux_decode.c:54
init
static av_cold int init(AVFilterContext *ctx)
Definition: vf_elbg.c:71
elbg.h
av_realloc_f
#define av_realloc_f(p, o, n)
Definition: tableprint_vlc.h:32
ELBGFilterContext::ctx
struct ELBGContext * ctx
Definition: vf_elbg.c:39
AVClass
Describe the class of an AVClass context structure.
Definition: log.h:66
NULL
#define NULL
Definition: coverity.c:32
av_frame_copy_props
int av_frame_copy_props(AVFrame *dst, const AVFrame *src)
Copy only "metadata" fields from src to dst.
Definition: frame.c:736
AVPixFmtDescriptor::nb_components
uint8_t nb_components
The number of components each pixel has, (1-4)
Definition: pixdesc.h:71
AV_PIX_FMT_ABGR
@ AV_PIX_FMT_ABGR
packed ABGR 8:8:8:8, 32bpp, ABGRABGR...
Definition: pixfmt.h:94
AVLFG
Context structure for the Lagged Fibonacci PRNG.
Definition: lfg.h:33
AV_PIX_FMT_RGB24
@ AV_PIX_FMT_RGB24
packed RGB 8:8:8, 24bpp, RGBRGB...
Definition: pixfmt.h:68
NULL_IF_CONFIG_SMALL
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
Definition: internal.h:106
avpriv_elbg_free
av_cold void avpriv_elbg_free(ELBGContext **elbgp)
Free an ELBGContext and reset the pointer to it.
Definition: elbg.c:515
ELBGFilterContext::max_steps_nb
int max_steps_nb
Definition: vf_elbg.c:42
B
#define B
Definition: vf_elbg.c:143
NB_COMPONENTS
#define NB_COMPONENTS
Definition: vf_elbg.c:111
filter_frame
static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
Definition: vf_elbg.c:146
elbg_options
static const AVOption elbg_options[]
Definition: vf_elbg.c:57
internal.h
AV_PIX_FMT_ARGB
@ AV_PIX_FMT_ARGB
packed ARGB 8:8:8:8, 32bpp, ARGBARGB...
Definition: pixfmt.h:92
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:255
ELBGContext
ELBG internal data.
Definition: elbg.c:46
AVFilterPad::name
const char * name
Pad name.
Definition: internal.h:53
AVFilter
Filter definition.
Definition: avfilter.h:166
AV_PIX_FMT_PAL8
@ AV_PIX_FMT_PAL8
8 bits with AV_PIX_FMT_RGB32 palette
Definition: pixfmt.h:77
ELBGFilterContext
Definition: vf_elbg.c:37
ret
ret
Definition: filter_design.txt:187
R
#define R
Definition: vf_elbg.c:141
FLAGS
#define FLAGS
Definition: vf_elbg.c:55
ELBGFilterContext::codebook_length
int codebook_length
Definition: vf_elbg.c:47
random_seed.h
AV_PIX_FMT_NONE
@ AV_PIX_FMT_NONE
Definition: pixfmt.h:65
AV_OPT_TYPE_INT
@ AV_OPT_TYPE_INT
Definition: opt.h:225
avfilter.h
AVFilterContext
An instance of a filter.
Definition: avfilter.h:397
AVMEDIA_TYPE_VIDEO
@ AVMEDIA_TYPE_VIDEO
Definition: avutil.h:201
uninit
static av_cold void uninit(AVFilterContext *ctx)
Definition: vf_elbg.c:236
AVFILTER_DEFINE_CLASS
AVFILTER_DEFINE_CLASS(elbg)
AVPixFmtDescriptor
Descriptor that unambiguously describes how the bits of a pixel are stored in the up to 4 data planes...
Definition: pixdesc.h:69
config_input
static int config_input(AVFilterLink *inlink)
Definition: vf_elbg.c:113
AV_OPT_TYPE_BOOL
@ AV_OPT_TYPE_BOOL
Definition: opt.h:244
FILTER_OUTPUTS
#define FILTER_OUTPUTS(array)
Definition: internal.h:193
av_freep
#define av_freep(p)
Definition: tableprint_vlc.h:34
ff_fill_rgba_map
int ff_fill_rgba_map(uint8_t *rgba_map, enum AVPixelFormat pix_fmt)
Definition: drawutils.c:35
ELBGFilterContext::codeword
int * codeword
Definition: vf_elbg.c:43
AVFrame::linesize
int linesize[AV_NUM_DATA_POINTERS]
For video, a positive or negative value, which is typically indicating the size in bytes of each pict...
Definition: frame.h:385
av_log
#define av_log(a,...)
Definition: tableprint_vlc.h:27
ELBGFilterContext::pal8
int pal8
Definition: vf_elbg.c:51
drawutils.h
AVFILTERPAD_FLAG_NEEDS_WRITABLE
#define AVFILTERPAD_FLAG_NEEDS_WRITABLE
The filter expects writable frames from its input link, duplicating data buffers if needed.
Definition: internal.h:66