Browse code

Implement cropdetect filter.

Originally committed as revision 25447 to svn://svn.ffmpeg.org/ffmpeg/trunk

Stefano Sabatini authored on 2010/10/12 18:22:32
Showing 7 changed files
... ...
@@ -45,6 +45,7 @@ version <next>:
45 45
 - Demuxer for Leitch/Harris' VR native stream format (LXF)
46 46
 - RTP depacketization of the X-QT QuickTime format
47 47
 - SAP (Session Announcement Protocol, RFC 2974) muxer
48
+- cropdetect filter
48 49
 
49 50
 
50 51
 version 0.6:
... ...
@@ -1399,6 +1399,7 @@ udp_protocol_deps="network"
1399 1399
 
1400 1400
 # filters
1401 1401
 blackframe_filter_deps="gpl"
1402
+cropdetect_filter_deps="gpl"
1402 1403
 frei0r_filter_deps="frei0r dlopen strtok_r"
1403 1404
 ocv_smooth_filter_deps="libopencv"
1404 1405
 yadif_filter_deps="gpl"
... ...
@@ -190,6 +190,41 @@ crop=in_w-100:in_h-100:100:100
190 190
 "crop=in_w/2:in_h/2:y:10+10*sin(n/10)"
191 191
 @end example
192 192
 
193
+@section cropdetect
194
+
195
+Auto-detect crop size.
196
+
197
+Calculate necessary cropping parameters and prints the recommended
198
+parameters through the logging system. The detected dimensions
199
+correspond to the non-black area of the input video.
200
+
201
+It accepts the syntax:
202
+@example
203
+cropdetect[=@var{limit}:@var{round}[:@var{reset}]]
204
+@end example
205
+
206
+@table @option
207
+
208
+@item limit
209
+Threshold, which can be optionally specified from nothing (0) to
210
+everything (255), defaults to 24.
211
+
212
+@item round
213
+Value which the width/height should be divisible by, defaults to
214
+16. The offset is automatically adjusted to center the video. Use 2 to
215
+get only even dimensions (needed for 4:2:2 video). 16 is best when
216
+encoding to most video codecs.
217
+
218
+@item reset
219
+Counter that determines after how many frames cropdetect will reset
220
+the previously detected largest video area and start over to detect
221
+the current optimal crop area. Defaults to 0.
222
+
223
+This can be useful when channel logos distort the video area. 0
224
+indicates never reset and return the largest area encountered during
225
+playback.
226
+@end table
227
+
193 228
 @section drawbox
194 229
 
195 230
 Draw a colored box on the input image.
... ...
@@ -23,6 +23,7 @@ OBJS-$(CONFIG_ANULLSINK_FILTER)              += asink_anullsink.o
23 23
 OBJS-$(CONFIG_ASPECT_FILTER)                 += vf_aspect.o
24 24
 OBJS-$(CONFIG_BLACKFRAME_FILTER)             += vf_blackframe.o
25 25
 OBJS-$(CONFIG_CROP_FILTER)                   += vf_crop.o
26
+OBJS-$(CONFIG_CROPDETECT_FILTER)             += vf_cropdetect.o
26 27
 OBJS-$(CONFIG_DRAWBOX_FILTER)                += vf_drawbox.o
27 28
 OBJS-$(CONFIG_FIFO_FILTER)                   += vf_fifo.o
28 29
 OBJS-$(CONFIG_FORMAT_FILTER)                 += vf_format.o
... ...
@@ -43,6 +43,7 @@ void avfilter_register_all(void)
43 43
     REGISTER_FILTER (ASPECT,      aspect,      vf);
44 44
     REGISTER_FILTER (BLACKFRAME,  blackframe,  vf);
45 45
     REGISTER_FILTER (CROP,        crop,        vf);
46
+    REGISTER_FILTER (CROPDETECT,  cropdetect,  vf);
46 47
     REGISTER_FILTER (DRAWBOX,     drawbox,     vf);
47 48
     REGISTER_FILTER (FIFO,        fifo,        vf);
48 49
     REGISTER_FILTER (FORMAT,      format,      vf);
... ...
@@ -25,8 +25,8 @@
25 25
 #include "libavutil/avutil.h"
26 26
 
27 27
 #define LIBAVFILTER_VERSION_MAJOR  1
28
-#define LIBAVFILTER_VERSION_MINOR 50
29
-#define LIBAVFILTER_VERSION_MICRO  1
28
+#define LIBAVFILTER_VERSION_MINOR 51
29
+#define LIBAVFILTER_VERSION_MICRO  0
30 30
 
31 31
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
32 32
                                                LIBAVFILTER_VERSION_MINOR, \
33 33
new file mode 100644
... ...
@@ -0,0 +1,214 @@
0
+/*
1
+ * Copyright (C) 2002 A'rpi
2
+ * This file is part of FFmpeg.
3
+ *
4
+ * FFmpeg is free software; you can redistribute it and/or modify
5
+ * it under the terms of the GNU General Public License as published by
6
+ * the Free Software Foundation; either version 2 of the License, or
7
+ * (at your option) any later version.
8
+ *
9
+ * FFmpeg is distributed in the hope that it will be useful,
10
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
11
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
12
+ * GNU General Public License for more details.
13
+ *
14
+ * You should have received a copy of the GNU General Public License along
15
+ * with FFmpeg; if not, write to the Free Software Foundation, Inc.,
16
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
17
+ */
18
+
19
+/**
20
+ * @file
21
+ * border detection filter
22
+ * Ported from MPlayer libmpcodecs/vf_cropdetect.c.
23
+ */
24
+
25
+#include "libavcore/imgutils.h"
26
+#include "avfilter.h"
27
+
28
+typedef struct {
29
+    int x1, y1, x2, y2;
30
+    int limit;
31
+    int round;
32
+    int reset_count;
33
+    int frame_nb;
34
+    int max_pixsteps[4];
35
+} CropDetectContext;
36
+
37
+static int query_formats(AVFilterContext *ctx)
38
+{
39
+    static const enum PixelFormat pix_fmts[] = {
40
+        PIX_FMT_YUV420P, PIX_FMT_YUVJ420P,
41
+        PIX_FMT_YUV422P, PIX_FMT_YUVJ422P,
42
+        PIX_FMT_YUV444P, PIX_FMT_YUVJ444P,
43
+        PIX_FMT_YUV411P, PIX_FMT_GRAY8,
44
+        PIX_FMT_NV12,    PIX_FMT_NV21,
45
+        PIX_FMT_NONE
46
+    };
47
+
48
+    avfilter_set_common_formats(ctx, avfilter_make_format_list(pix_fmts));
49
+    return 0;
50
+}
51
+
52
+static int checkline(void *ctx, const unsigned char *src, int stride, int len, int bpp)
53
+{
54
+    int total = 0;
55
+    int div = len;
56
+
57
+    switch (bpp) {
58
+    case 1:
59
+        while (--len >= 0) {
60
+            total += src[0];
61
+            src += stride;
62
+        }
63
+        break;
64
+    case 3:
65
+    case 4:
66
+        while (--len >= 0) {
67
+            total += src[0] + src[1] + src[2];
68
+            src += stride;
69
+        }
70
+        div *= 3;
71
+        break;
72
+    }
73
+    total /= div;
74
+
75
+    av_log(ctx, AV_LOG_DEBUG, "total:%d\n", total);
76
+    return total;
77
+}
78
+
79
+static av_cold int init(AVFilterContext *ctx, const char *args, void *opaque)
80
+{
81
+    CropDetectContext *cd = ctx->priv;
82
+
83
+    cd->limit = 24;
84
+    cd->round = 0;
85
+    cd->reset_count = 0;
86
+    cd->frame_nb = -2;
87
+
88
+    if (args)
89
+        sscanf(args, "%d:%d:%d", &cd->limit, &cd->round, &cd->reset_count);
90
+
91
+    av_log(ctx, AV_LOG_INFO, "limit:%d round:%d reset_count:%d\n",
92
+           cd->limit, cd->round, cd->reset_count);
93
+
94
+    return 0;
95
+}
96
+
97
+static int config_input(AVFilterLink *inlink)
98
+{
99
+    AVFilterContext *ctx = inlink->dst;
100
+    CropDetectContext *cd = ctx->priv;
101
+
102
+    av_image_fill_max_pixsteps(cd->max_pixsteps, NULL,
103
+                               &av_pix_fmt_descriptors[inlink->format]);
104
+
105
+    cd->x1 = inlink->w - 1;
106
+    cd->y1 = inlink->h - 1;
107
+    cd->x2 = 0;
108
+    cd->y2 = 0;
109
+
110
+    return 0;
111
+}
112
+
113
+static void end_frame(AVFilterLink *inlink)
114
+{
115
+    AVFilterContext *ctx = inlink->dst;
116
+    CropDetectContext *cd = ctx->priv;
117
+    AVFilterBufferRef *picref = inlink->cur_buf;
118
+    int bpp = cd->max_pixsteps[0];
119
+    int w, h, x, y, shrink_by;
120
+
121
+    // ignore first 2 frames - they may be empty
122
+    if (++cd->frame_nb > 0) {
123
+        // Reset the crop area every reset_count frames, if reset_count is > 0
124
+        if (cd->reset_count > 0 && cd->frame_nb > cd->reset_count) {
125
+            cd->x1 = picref->video->w-1;
126
+            cd->y1 = picref->video->h-1;
127
+            cd->x2 = 0;
128
+            cd->y2 = 0;
129
+            cd->frame_nb = 1;
130
+        }
131
+
132
+        for (y = 0; y < cd->y1; y++) {
133
+            if (checkline(ctx, picref->data[0] + picref->linesize[0] * y, bpp, picref->video->w, bpp) > cd->limit) {
134
+                cd->y1 = y;
135
+                break;
136
+            }
137
+        }
138
+
139
+        for (y = picref->video->h-1; y > cd->y2; y--) {
140
+            if (checkline(ctx, picref->data[0] + picref->linesize[0] * y, bpp, picref->video->w, bpp) > cd->limit) {
141
+                cd->y2 = y;
142
+                break;
143
+            }
144
+        }
145
+
146
+        for (y = 0; y < cd->x1; y++) {
147
+            if (checkline(ctx, picref->data[0] + bpp*y, picref->linesize[0], picref->video->h, bpp) > cd->limit) {
148
+                cd->x1 = y;
149
+                break;
150
+            }
151
+        }
152
+
153
+        for (y = picref->video->w-1; y > cd->x2; y--) {
154
+            if (checkline(ctx, picref->data[0] + bpp*y, picref->linesize[0], picref->video->h, bpp) > cd->limit) {
155
+                cd->x2 = y;
156
+                break;
157
+            }
158
+        }
159
+
160
+        // round x and y (up), important for yuv colorspaces
161
+        // make sure they stay rounded!
162
+        x = (cd->x1+1) & ~1;
163
+        y = (cd->y1+1) & ~1;
164
+
165
+        w = cd->x2 - x + 1;
166
+        h = cd->y2 - y + 1;
167
+
168
+        // w and h must be divisible by 2 as well because of yuv
169
+        // colorspace problems.
170
+        if (cd->round <= 1)
171
+            cd->round = 16;
172
+        if (cd->round % 2)
173
+            cd->round *= 2;
174
+
175
+        shrink_by = w % cd->round;
176
+        w -= shrink_by;
177
+        x += (shrink_by/2 + 1) & ~1;
178
+
179
+        shrink_by = h % cd->round;
180
+        h -= shrink_by;
181
+        y += (shrink_by/2 + 1) & ~1;
182
+
183
+        av_log(ctx, AV_LOG_INFO,
184
+               "x1:%d x2:%d y1:%d y2:%d w:%d h:%d x:%d y:%d pos:%"PRId64" pts:%f crop=%d:%d:%d:%d\n",
185
+               cd->x1, cd->x2, cd->y1, cd->y2, w, h, x, y, picref->pos,
186
+               picref->pts == AV_NOPTS_VALUE ? -1 : (double)picref->pts / AV_TIME_BASE,
187
+               w, h, x, y);
188
+    }
189
+
190
+    avfilter_end_frame(inlink->dst->outputs[0]);
191
+}
192
+
193
+AVFilter avfilter_vf_cropdetect = {
194
+    .name        = "cropdetect",
195
+    .description = NULL_IF_CONFIG_SMALL("Auto-detect crop size."),
196
+
197
+    .priv_size = sizeof(CropDetectContext),
198
+    .init      = init,
199
+
200
+    .query_formats = query_formats,
201
+
202
+    .inputs    = (AVFilterPad[]) {{ .name = "default",
203
+                                    .type             = AVMEDIA_TYPE_VIDEO,
204
+                                    .config_props     = config_input,
205
+                                    .get_video_buffer = avfilter_null_get_video_buffer,
206
+                                    .start_frame      = avfilter_null_start_frame,
207
+                                    .end_frame        = end_frame, },
208
+                                  { .name = NULL}},
209
+
210
+    .outputs   = (AVFilterPad[]) {{ .name             = "default",
211
+                                    .type             = AVMEDIA_TYPE_VIDEO },
212
+                                  { .name = NULL}},
213
+};