FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
vf_delogo.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2002 Jindrich Makovicka <makovick@gmail.com>
3  * Copyright (c) 2011 Stefano Sabatini
4  * Copyright (c) 2013, 2015 Jean Delvare <jdelvare@suse.com>
5  *
6  * This file is part of FFmpeg.
7  *
8  * FFmpeg is free software; you can redistribute it and/or modify
9  * it under the terms of the GNU General Public License as published by
10  * the Free Software Foundation; either version 2 of the License, or
11  * (at your option) any later version.
12  *
13  * FFmpeg is distributed in the hope that it will be useful,
14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16  * GNU General Public License for more details.
17  *
18  * You should have received a copy of the GNU General Public License along
19  * with FFmpeg; if not, write to the Free Software Foundation, Inc.,
20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
21  */
22 
23 /**
24  * @file
25  * A very simple tv station logo remover
26  * Originally imported from MPlayer libmpcodecs/vf_delogo.c,
27  * the algorithm was later improved.
28  */
29 
30 #include "libavutil/common.h"
31 #include "libavutil/imgutils.h"
32 #include "libavutil/opt.h"
33 #include "libavutil/pixdesc.h"
34 #include "avfilter.h"
35 #include "formats.h"
36 #include "internal.h"
37 #include "video.h"
38 
39 /**
40  * Apply a simple delogo algorithm to the image in src and put the
41  * result in dst.
42  *
43  * The algorithm is only applied to the region specified by the logo
44  * parameters.
45  *
46  * @param w width of the input image
47  * @param h height of the input image
48  * @param logo_x x coordinate of the top left corner of the logo region
49  * @param logo_y y coordinate of the top left corner of the logo region
50  * @param logo_w width of the logo
51  * @param logo_h height of the logo
52  * @param band the size of the band around the processed area
53  * @param show show a rectangle around the processed area, useful for
54  * parameters tweaking
55  * @param direct if non-zero perform in-place processing
56  */
57 static void apply_delogo(uint8_t *dst, int dst_linesize,
58  uint8_t *src, int src_linesize,
59  int w, int h, AVRational sar,
60  int logo_x, int logo_y, int logo_w, int logo_h,
61  unsigned int band, int show, int direct)
62 {
63  int x, y;
64  uint64_t interp, weightl, weightr, weightt, weightb, weight;
65  uint8_t *xdst, *xsrc;
66 
67  uint8_t *topleft, *botleft, *topright;
68  unsigned int left_sample, right_sample;
69  int xclipl, xclipr, yclipt, yclipb;
70  int logo_x1, logo_x2, logo_y1, logo_y2;
71 
72  xclipl = FFMAX(-logo_x, 0);
73  xclipr = FFMAX(logo_x+logo_w-w, 0);
74  yclipt = FFMAX(-logo_y, 0);
75  yclipb = FFMAX(logo_y+logo_h-h, 0);
76 
77  logo_x1 = logo_x + xclipl;
78  logo_x2 = logo_x + logo_w - xclipr - 1;
79  logo_y1 = logo_y + yclipt;
80  logo_y2 = logo_y + logo_h - yclipb - 1;
81 
82  topleft = src+logo_y1 * src_linesize+logo_x1;
83  topright = src+logo_y1 * src_linesize+logo_x2;
84  botleft = src+logo_y2 * src_linesize+logo_x1;
85 
86  if (!direct)
87  av_image_copy_plane(dst, dst_linesize, src, src_linesize, w, h);
88 
89  dst += (logo_y1 + 1) * dst_linesize;
90  src += (logo_y1 + 1) * src_linesize;
91 
92  for (y = logo_y1+1; y < logo_y2; y++) {
93  left_sample = topleft[src_linesize*(y-logo_y1)] +
94  topleft[src_linesize*(y-logo_y1-1)] +
95  topleft[src_linesize*(y-logo_y1+1)];
96  right_sample = topright[src_linesize*(y-logo_y1)] +
97  topright[src_linesize*(y-logo_y1-1)] +
98  topright[src_linesize*(y-logo_y1+1)];
99 
100  for (x = logo_x1+1,
101  xdst = dst+logo_x1+1,
102  xsrc = src+logo_x1+1; x < logo_x2; x++, xdst++, xsrc++) {
103 
104  if (show && (y == logo_y1+1 || y == logo_y2-1 ||
105  x == logo_x1+1 || x == logo_x2-1)) {
106  *xdst = 0;
107  continue;
108  }
109 
110  /* Weighted interpolation based on relative distances, taking SAR into account */
111  weightl = (uint64_t) (logo_x2-x) * (y-logo_y1) * (logo_y2-y) * sar.den;
112  weightr = (uint64_t)(x-logo_x1) * (y-logo_y1) * (logo_y2-y) * sar.den;
113  weightt = (uint64_t)(x-logo_x1) * (logo_x2-x) * (logo_y2-y) * sar.num;
114  weightb = (uint64_t)(x-logo_x1) * (logo_x2-x) * (y-logo_y1) * sar.num;
115 
116  interp =
117  left_sample * weightl
118  +
119  right_sample * weightr
120  +
121  (topleft[x-logo_x1] +
122  topleft[x-logo_x1-1] +
123  topleft[x-logo_x1+1]) * weightt
124  +
125  (botleft[x-logo_x1] +
126  botleft[x-logo_x1-1] +
127  botleft[x-logo_x1+1]) * weightb;
128  weight = (weightl + weightr + weightt + weightb) * 3U;
129  interp = ROUNDED_DIV(interp, weight);
130 
131  if (y >= logo_y+band && y < logo_y+logo_h-band &&
132  x >= logo_x+band && x < logo_x+logo_w-band) {
133  *xdst = interp;
134  } else {
135  unsigned dist = 0;
136 
137  if (x < logo_x+band)
138  dist = FFMAX(dist, logo_x-x+band);
139  else if (x >= logo_x+logo_w-band)
140  dist = FFMAX(dist, x-(logo_x+logo_w-1-band));
141 
142  if (y < logo_y+band)
143  dist = FFMAX(dist, logo_y-y+band);
144  else if (y >= logo_y+logo_h-band)
145  dist = FFMAX(dist, y-(logo_y+logo_h-1-band));
146 
147  *xdst = (*xsrc*dist + interp*(band-dist))/band;
148  }
149  }
150 
151  dst += dst_linesize;
152  src += src_linesize;
153  }
154 }
155 
156 typedef struct DelogoContext {
157  const AVClass *class;
158  int x, y, w, h, band, show;
159 } DelogoContext;
160 
161 #define OFFSET(x) offsetof(DelogoContext, x)
162 #define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
163 
164 static const AVOption delogo_options[]= {
165  { "x", "set logo x position", OFFSET(x), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, INT_MAX, FLAGS },
166  { "y", "set logo y position", OFFSET(y), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, INT_MAX, FLAGS },
167  { "w", "set logo width", OFFSET(w), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, INT_MAX, FLAGS },
168  { "h", "set logo height", OFFSET(h), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, INT_MAX, FLAGS },
169 #if LIBAVFILTER_VERSION_MAJOR < 7
170  /* Actual default value for band/t is 1, set in init */
171  { "band", "set delogo area band size", OFFSET(band), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, FLAGS },
172  { "t", "set delogo area band size", OFFSET(band), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, FLAGS },
173 #endif
174  { "show", "show delogo area", OFFSET(show), AV_OPT_TYPE_BOOL,{ .i64 = 0 }, 0, 1, FLAGS },
175  { NULL }
176 };
177 
178 AVFILTER_DEFINE_CLASS(delogo);
179 
181 {
182  static const enum AVPixelFormat pix_fmts[] = {
187  };
188  AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts);
189  if (!fmts_list)
190  return AVERROR(ENOMEM);
191  return ff_set_common_formats(ctx, fmts_list);
192 }
193 
195 {
196  DelogoContext *s = ctx->priv;
197 
198 #define CHECK_UNSET_OPT(opt) \
199  if (s->opt == -1) { \
200  av_log(s, AV_LOG_ERROR, "Option %s was not set.\n", #opt); \
201  return AVERROR(EINVAL); \
202  }
203  CHECK_UNSET_OPT(x);
204  CHECK_UNSET_OPT(y);
205  CHECK_UNSET_OPT(w);
207 
208 #if LIBAVFILTER_VERSION_MAJOR < 7
209  if (s->band == 0) { /* Unset, use default */
210  av_log(ctx, AV_LOG_WARNING, "Note: default band value was changed from 4 to 1.\n");
211  s->band = 1;
212  } else if (s->band != 1) {
213  av_log(ctx, AV_LOG_WARNING, "Option band is deprecated.\n");
214  }
215 #else
216  s->band = 1;
217 #endif
218  av_log(ctx, AV_LOG_VERBOSE, "x:%d y:%d, w:%d h:%d band:%d show:%d\n",
219  s->x, s->y, s->w, s->h, s->band, s->show);
220 
221  s->w += s->band*2;
222  s->h += s->band*2;
223  s->x -= s->band;
224  s->y -= s->band;
225 
226  return 0;
227 }
228 
229 static int filter_frame(AVFilterLink *inlink, AVFrame *in)
230 {
231  DelogoContext *s = inlink->dst->priv;
232  AVFilterLink *outlink = inlink->dst->outputs[0];
233  const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
234  AVFrame *out;
235  int hsub0 = desc->log2_chroma_w;
236  int vsub0 = desc->log2_chroma_h;
237  int direct = 0;
238  int plane;
239  AVRational sar;
240 
241  if (av_frame_is_writable(in)) {
242  direct = 1;
243  out = in;
244  } else {
245  out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
246  if (!out) {
247  av_frame_free(&in);
248  return AVERROR(ENOMEM);
249  }
250 
251  av_frame_copy_props(out, in);
252  }
253 
254  sar = in->sample_aspect_ratio;
255  /* Assume square pixels if SAR is unknown */
256  if (!sar.num)
257  sar.num = sar.den = 1;
258 
259  for (plane = 0; plane < desc->nb_components; plane++) {
260  int hsub = plane == 1 || plane == 2 ? hsub0 : 0;
261  int vsub = plane == 1 || plane == 2 ? vsub0 : 0;
262 
263  apply_delogo(out->data[plane], out->linesize[plane],
264  in ->data[plane], in ->linesize[plane],
265  AV_CEIL_RSHIFT(inlink->w, hsub),
266  AV_CEIL_RSHIFT(inlink->h, vsub),
267  sar, s->x>>hsub, s->y>>vsub,
268  /* Up and left borders were rounded down, inject lost bits
269  * into width and height to avoid error accumulation */
270  AV_CEIL_RSHIFT(s->w + (s->x & ((1<<hsub)-1)), hsub),
271  AV_CEIL_RSHIFT(s->h + (s->y & ((1<<vsub)-1)), vsub),
272  s->band>>FFMIN(hsub, vsub),
273  s->show, direct);
274  }
275 
276  if (!direct)
277  av_frame_free(&in);
278 
279  return ff_filter_frame(outlink, out);
280 }
281 
283  {
284  .name = "default",
285  .type = AVMEDIA_TYPE_VIDEO,
286  .filter_frame = filter_frame,
287  },
288  { NULL }
289 };
290 
292  {
293  .name = "default",
294  .type = AVMEDIA_TYPE_VIDEO,
295  },
296  { NULL }
297 };
298 
300  .name = "delogo",
301  .description = NULL_IF_CONFIG_SMALL("Remove logo from input video."),
302  .priv_size = sizeof(DelogoContext),
303  .priv_class = &delogo_class,
304  .init = init,
306  .inputs = avfilter_vf_delogo_inputs,
307  .outputs = avfilter_vf_delogo_outputs,
309 };
int plane
Definition: avisynth_c.h:291
#define NULL
Definition: coverity.c:32
const char * s
Definition: avisynth_c.h:631
const AVPixFmtDescriptor * av_pix_fmt_desc_get(enum AVPixelFormat pix_fmt)
Definition: pixdesc.c:2157
This structure describes decoded (raw) audio or video data.
Definition: frame.h:181
AVOption.
Definition: opt.h:245
AVFormatContext * ctx
Definition: movenc-test.c:48
planar YUV 4:4:4, 24bpp, (1 Cr & Cb sample per 1x1 Y samples)
Definition: pixfmt.h:68
misc image utilities
#define AV_LOG_WARNING
Something somehow does not look correct.
Definition: log.h:182
Main libavfilter public API header.
static void apply_delogo(uint8_t *dst, int dst_linesize, uint8_t *src, int src_linesize, int w, int h, AVRational sar, int logo_x, int logo_y, int logo_w, int logo_h, unsigned int band, int show, int direct)
Apply a simple delogo algorithm to the image in src and put the result in dst.
Definition: vf_delogo.c:57
#define FLAGS
Definition: vf_delogo.c:162
int num
numerator
Definition: rational.h:44
static int query_formats(AVFilterContext *ctx)
Definition: vf_delogo.c:180
AVFrame * ff_get_video_buffer(AVFilterLink *link, int w, int h)
Request a picture buffer with a specific set of permissions.
Definition: video.c:76
uint8_t log2_chroma_w
Amount to shift the luma width right to find the chroma width.
Definition: pixdesc.h:92
AVFilterFormats * ff_make_format_list(const int *fmts)
Create a list of supported formats.
Definition: formats.c:283
#define AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC
Some filters support a generic "enable" expression option that can be used to enable or disable a fil...
Definition: avfilter.h:122
const char * name
Pad name.
Definition: internal.h:59
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:1163
planar YUV 4:2:0, 20bpp, (1 Cr & Cb sample per 2x2 Y & A samples)
Definition: pixfmt.h:103
uint8_t
#define av_cold
Definition: attributes.h:82
AVOptions.
#define AV_LOG_VERBOSE
Detailed information.
Definition: log.h:192
AVFilter ff_vf_delogo
Definition: vf_delogo.c:299
#define av_log(a,...)
#define ROUNDED_DIV(a, b)
Definition: common.h:56
A filter pad used for either input or output.
Definition: internal.h:53
#define U(x)
Definition: vp56_arith.h:37
int ff_set_common_formats(AVFilterContext *ctx, AVFilterFormats *formats)
A helper for query_formats() which sets all links to the same list of formats.
Definition: formats.c:568
uint8_t log2_chroma_h
Amount to shift the luma height right to find the chroma height.
Definition: pixdesc.h:101
static const AVFilterPad avfilter_vf_delogo_outputs[]
Definition: vf_delogo.c:291
#define AVERROR(e)
Definition: error.h:43
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:154
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification. ...
Definition: internal.h:176
void * priv
private data for use by the filter
Definition: avfilter.h:319
static av_cold int init(AVFilterContext *ctx)
Definition: vf_delogo.c:194
#define OFFSET(x)
Definition: vf_delogo.c:161
#define FFMAX(a, b)
Definition: common.h:94
planar YUV 4:2:2, 16bpp, (1 Cr & Cb sample per 2x1 Y samples)
Definition: pixfmt.h:67
uint8_t nb_components
The number of components each pixel has, (1-4)
Definition: pixdesc.h:83
#define FFMIN(a, b)
Definition: common.h:96
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
Definition: vf_delogo.c:229
static const AVFilterPad outputs[]
Definition: af_afftfilt.c:385
#define src
Definition: vp9dsp.c:530
FILE * out
Definition: movenc-test.c:54
AVFILTER_DEFINE_CLASS(delogo)
static const AVFilterPad inputs[]
Definition: af_afftfilt.c:375
int av_frame_is_writable(AVFrame *frame)
Check if the frame data is writable.
Definition: frame.c:505
int linesize[AV_NUM_DATA_POINTERS]
For video, size in bytes of each picture line.
Definition: frame.h:209
Descriptor that unambiguously describes how the bits of a pixel are stored in the up to 4 data planes...
Definition: pixdesc.h:81
AVRational sample_aspect_ratio
Sample aspect ratio for the video frame, 0/1 if unknown/unspecified.
Definition: frame.h:257
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(constint16_t *) pi >>8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(constint32_t *) pi >>24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(constfloat *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(constfloat *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(constfloat *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(constdouble *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(constdouble *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(constdouble *) pi *(1U<< 31))))#defineSET_CONV_FUNC_GROUP(ofmt, ifmt) staticvoidset_generic_function(AudioConvert *ac){}voidff_audio_convert_free(AudioConvert **ac){if(!*ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);}AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enumAVSampleFormatout_fmt, enumAVSampleFormatin_fmt, intchannels, intsample_rate, intapply_map){AudioConvert *ac;intin_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) returnNULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method!=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt)>2){ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc){av_free(ac);returnNULL;}returnac;}in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar){ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar?ac->channels:1;}elseif(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;elseac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);returnac;}intff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in){intuse_generic=1;intlen=in->nb_samples;intp;if(ac->dc){av_log(ac->avr, AV_LOG_TRACE,"%dsamples-audio_convert:%sto%s(dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));returnff_convert_dither(ac-> in
static const AVOption delogo_options[]
Definition: vf_delogo.c:164
planar YUV 4:1:0, 9bpp, (1 Cr & Cb sample per 4x4 Y samples)
Definition: pixfmt.h:69
Describe the class of an AVClass context structure.
Definition: log.h:67
Filter definition.
Definition: avfilter.h:141
rational number numerator/denominator
Definition: rational.h:43
const char * name
Filter name.
Definition: avfilter.h:145
AVFilterLink ** outputs
array of pointers to output links
Definition: avfilter.h:316
static int weight(int i, int blen, int offset)
Definition: diracdec.c:1413
static enum AVPixelFormat pix_fmts[]
Definition: libkvazaar.c:262
static int flags
Definition: cpu.c:47
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:192
#define CHECK_UNSET_OPT(opt)
planar YUV 4:2:0, 12bpp, (1 Cr & Cb sample per 2x2 Y samples)
Definition: pixfmt.h:63
Y , 8bpp.
Definition: pixfmt.h:71
common internal and external API header
planar YUV 4:1:1, 12bpp, (1 Cr & Cb sample per 4x1 Y samples)
Definition: pixfmt.h:70
int den
denominator
Definition: rational.h:45
A list of supported formats for one end of a filter link.
Definition: formats.h:64
An instance of a filter.
Definition: avfilter.h:304
planar YUV 4:4:0 (1 Cr & Cb sample per 1x2 Y samples)
Definition: pixfmt.h:101
void av_image_copy_plane(uint8_t *dst, int dst_linesize, const uint8_t *src, int src_linesize, int bytewidth, int height)
Copy image plane from src to dst.
Definition: imgutils.c:287
internal API functions
static const AVFilterPad avfilter_vf_delogo_inputs[]
Definition: vf_delogo.c:282
AVPixelFormat
Pixel format.
Definition: pixfmt.h:61
int av_frame_copy_props(AVFrame *dst, const AVFrame *src)
Copy only "metadata" fields from src to dst.
Definition: frame.c:565
#define AV_CEIL_RSHIFT(a, b)
Definition: common.h:58