blob: 47429494f3f9932723d3d19a36487c58e3b8c5cd
1 | /* |
2 | * Copyright (c) 2013 Georg Martius <georg dot martius at web dot de> |
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 | #define DEFAULT_RESULT_NAME "transforms.trf" |
22 | |
23 | #include <vid.stab/libvidstab.h> |
24 | |
25 | #include "libavutil/common.h" |
26 | #include "libavutil/opt.h" |
27 | #include "libavutil/imgutils.h" |
28 | #include "avfilter.h" |
29 | #include "internal.h" |
30 | |
31 | #include "vidstabutils.h" |
32 | |
33 | typedef struct { |
34 | const AVClass *class; |
35 | |
36 | VSMotionDetect md; |
37 | VSMotionDetectConfig conf; |
38 | |
39 | char *result; |
40 | FILE *f; |
41 | } StabData; |
42 | |
43 | |
44 | #define OFFSET(x) offsetof(StabData, x) |
45 | #define OFFSETC(x) (offsetof(StabData, conf)+offsetof(VSMotionDetectConfig, x)) |
46 | #define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM |
47 | |
48 | static const AVOption vidstabdetect_options[] = { |
49 | {"result", "path to the file used to write the transforms", OFFSET(result), AV_OPT_TYPE_STRING, {.str = DEFAULT_RESULT_NAME}, .flags = FLAGS}, |
50 | {"shakiness", "how shaky is the video and how quick is the camera?" |
51 | " 1: little (fast) 10: very strong/quick (slow)", OFFSETC(shakiness), AV_OPT_TYPE_INT, {.i64 = 5}, 1, 10, FLAGS}, |
52 | {"accuracy", "(>=shakiness) 1: low 15: high (slow)", OFFSETC(accuracy), AV_OPT_TYPE_INT, {.i64 = 15}, 1, 15, FLAGS}, |
53 | {"stepsize", "region around minimum is scanned with 1 pixel resolution", OFFSETC(stepSize), AV_OPT_TYPE_INT, {.i64 = 6}, 1, 32, FLAGS}, |
54 | {"mincontrast", "below this contrast a field is discarded (0-1)", OFFSETC(contrastThreshold), AV_OPT_TYPE_DOUBLE, {.dbl = 0.25}, 0.0, 1.0, FLAGS}, |
55 | {"show", "0: draw nothing; 1,2: show fields and transforms", OFFSETC(show), AV_OPT_TYPE_INT, {.i64 = 0}, 0, 2, FLAGS}, |
56 | {"tripod", "virtual tripod mode (if >0): motion is compared to a reference" |
57 | " reference frame (frame # is the value)", OFFSETC(virtualTripod), AV_OPT_TYPE_INT, {.i64 = 0}, 0, INT_MAX, FLAGS}, |
58 | {NULL} |
59 | }; |
60 | |
61 | AVFILTER_DEFINE_CLASS(vidstabdetect); |
62 | |
63 | static av_cold int init(AVFilterContext *ctx) |
64 | { |
65 | StabData *s = ctx->priv; |
66 | ff_vs_init(); |
67 | s->class = &vidstabdetect_class; |
68 | av_log(ctx, AV_LOG_VERBOSE, "vidstabdetect filter: init %s\n", LIBVIDSTAB_VERSION); |
69 | return 0; |
70 | } |
71 | |
72 | static av_cold void uninit(AVFilterContext *ctx) |
73 | { |
74 | StabData *s = ctx->priv; |
75 | VSMotionDetect *md = &(s->md); |
76 | |
77 | if (s->f) { |
78 | fclose(s->f); |
79 | s->f = NULL; |
80 | } |
81 | |
82 | vsMotionDetectionCleanup(md); |
83 | } |
84 | |
85 | static int query_formats(AVFilterContext *ctx) |
86 | { |
87 | // If you add something here also add it in vidstabutils.c |
88 | static const enum AVPixelFormat pix_fmts[] = { |
89 | AV_PIX_FMT_YUV444P, AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUV420P, |
90 | AV_PIX_FMT_YUV411P, AV_PIX_FMT_YUV410P, AV_PIX_FMT_YUVA420P, |
91 | AV_PIX_FMT_YUV440P, AV_PIX_FMT_GRAY8, |
92 | AV_PIX_FMT_RGB24, AV_PIX_FMT_BGR24, AV_PIX_FMT_RGBA, |
93 | AV_PIX_FMT_NONE |
94 | }; |
95 | |
96 | AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts); |
97 | if (!fmts_list) |
98 | return AVERROR(ENOMEM); |
99 | return ff_set_common_formats(ctx, fmts_list); |
100 | } |
101 | |
102 | static int config_input(AVFilterLink *inlink) |
103 | { |
104 | AVFilterContext *ctx = inlink->dst; |
105 | StabData *s = ctx->priv; |
106 | |
107 | VSMotionDetect* md = &(s->md); |
108 | VSFrameInfo fi; |
109 | const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format); |
110 | |
111 | vsFrameInfoInit(&fi, inlink->w, inlink->h, |
112 | ff_av2vs_pixfmt(ctx, inlink->format)); |
113 | if (fi.bytesPerPixel != av_get_bits_per_pixel(desc)/8) { |
114 | av_log(ctx, AV_LOG_ERROR, "pixel-format error: wrong bits/per/pixel, please report a BUG"); |
115 | return AVERROR(EINVAL); |
116 | } |
117 | if (fi.log2ChromaW != desc->log2_chroma_w) { |
118 | av_log(ctx, AV_LOG_ERROR, "pixel-format error: log2_chroma_w, please report a BUG"); |
119 | return AVERROR(EINVAL); |
120 | } |
121 | |
122 | if (fi.log2ChromaH != desc->log2_chroma_h) { |
123 | av_log(ctx, AV_LOG_ERROR, "pixel-format error: log2_chroma_h, please report a BUG"); |
124 | return AVERROR(EINVAL); |
125 | } |
126 | |
127 | // set values that are not initialized by the options |
128 | s->conf.algo = 1; |
129 | s->conf.modName = "vidstabdetect"; |
130 | if (vsMotionDetectInit(md, &s->conf, &fi) != VS_OK) { |
131 | av_log(ctx, AV_LOG_ERROR, "initialization of Motion Detection failed, please report a BUG"); |
132 | return AVERROR(EINVAL); |
133 | } |
134 | |
135 | vsMotionDetectGetConfig(&s->conf, md); |
136 | av_log(ctx, AV_LOG_INFO, "Video stabilization settings (pass 1/2):\n"); |
137 | av_log(ctx, AV_LOG_INFO, " shakiness = %d\n", s->conf.shakiness); |
138 | av_log(ctx, AV_LOG_INFO, " accuracy = %d\n", s->conf.accuracy); |
139 | av_log(ctx, AV_LOG_INFO, " stepsize = %d\n", s->conf.stepSize); |
140 | av_log(ctx, AV_LOG_INFO, " mincontrast = %f\n", s->conf.contrastThreshold); |
141 | av_log(ctx, AV_LOG_INFO, " tripod = %d\n", s->conf.virtualTripod); |
142 | av_log(ctx, AV_LOG_INFO, " show = %d\n", s->conf.show); |
143 | av_log(ctx, AV_LOG_INFO, " result = %s\n", s->result); |
144 | |
145 | s->f = fopen(s->result, "w"); |
146 | if (s->f == NULL) { |
147 | av_log(ctx, AV_LOG_ERROR, "cannot open transform file %s\n", s->result); |
148 | return AVERROR(EINVAL); |
149 | } else { |
150 | if (vsPrepareFile(md, s->f) != VS_OK) { |
151 | av_log(ctx, AV_LOG_ERROR, "cannot write to transform file %s\n", s->result); |
152 | return AVERROR(EINVAL); |
153 | } |
154 | } |
155 | return 0; |
156 | } |
157 | |
158 | static int filter_frame(AVFilterLink *inlink, AVFrame *in) |
159 | { |
160 | AVFilterContext *ctx = inlink->dst; |
161 | StabData *s = ctx->priv; |
162 | VSMotionDetect *md = &(s->md); |
163 | LocalMotions localmotions; |
164 | |
165 | AVFilterLink *outlink = inlink->dst->outputs[0]; |
166 | VSFrame frame; |
167 | int plane; |
168 | |
169 | if (s->conf.show > 0 && !av_frame_is_writable(in)) |
170 | av_frame_make_writable(in); |
171 | |
172 | for (plane = 0; plane < md->fi.planes; plane++) { |
173 | frame.data[plane] = in->data[plane]; |
174 | frame.linesize[plane] = in->linesize[plane]; |
175 | } |
176 | if (vsMotionDetection(md, &localmotions, &frame) != VS_OK) { |
177 | av_log(ctx, AV_LOG_ERROR, "motion detection failed"); |
178 | return AVERROR(AVERROR_EXTERNAL); |
179 | } else { |
180 | if (vsWriteToFile(md, s->f, &localmotions) != VS_OK) { |
181 | int ret = AVERROR(errno); |
182 | av_log(ctx, AV_LOG_ERROR, "cannot write to transform file"); |
183 | return ret; |
184 | } |
185 | vs_vector_del(&localmotions); |
186 | } |
187 | |
188 | return ff_filter_frame(outlink, in); |
189 | } |
190 | |
191 | static const AVFilterPad avfilter_vf_vidstabdetect_inputs[] = { |
192 | { |
193 | .name = "default", |
194 | .type = AVMEDIA_TYPE_VIDEO, |
195 | .filter_frame = filter_frame, |
196 | .config_props = config_input, |
197 | }, |
198 | { NULL } |
199 | }; |
200 | |
201 | static const AVFilterPad avfilter_vf_vidstabdetect_outputs[] = { |
202 | { |
203 | .name = "default", |
204 | .type = AVMEDIA_TYPE_VIDEO, |
205 | }, |
206 | { NULL } |
207 | }; |
208 | |
209 | AVFilter ff_vf_vidstabdetect = { |
210 | .name = "vidstabdetect", |
211 | .description = NULL_IF_CONFIG_SMALL("Extract relative transformations, " |
212 | "pass 1 of 2 for stabilization " |
213 | "(see vidstabtransform for pass 2)."), |
214 | .priv_size = sizeof(StabData), |
215 | .init = init, |
216 | .uninit = uninit, |
217 | .query_formats = query_formats, |
218 | .inputs = avfilter_vf_vidstabdetect_inputs, |
219 | .outputs = avfilter_vf_vidstabdetect_outputs, |
220 | .priv_class = &vidstabdetect_class, |
221 | }; |
222 |