ffmpeg/libavfilter/vf_crop.c

410 lines
14 KiB
C
Raw Permalink Normal View History

/*
* Copyright (c) 2007 Bobby Bingham
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* video crop filter
*/
#include <stdio.h>
#include "avfilter.h"
#include "formats.h"
#include "internal.h"
#include "video.h"
#include "libavutil/eval.h"
#include "libavutil/avstring.h"
#include "libavutil/internal.h"
#include "libavutil/libm.h"
#include "libavutil/imgutils.h"
#include "libavutil/mathematics.h"
#include "libavutil/opt.h"
static const char *const var_names[] = {
"in_w", "iw", ///< width of the input video
"in_h", "ih", ///< height of the input video
"out_w", "ow", ///< width of the cropped video
"out_h", "oh", ///< height of the cropped video
"a",
"sar",
"dar",
"hsub",
"vsub",
"x",
"y",
"n", ///< number of frame
#if FF_API_FRAME_PKT
"pos", ///< position in the file
#endif
"t", ///< timestamp expressed in seconds
NULL
};
enum var_name {
VAR_IN_W, VAR_IW,
VAR_IN_H, VAR_IH,
VAR_OUT_W, VAR_OW,
VAR_OUT_H, VAR_OH,
VAR_A,
VAR_SAR,
VAR_DAR,
VAR_HSUB,
VAR_VSUB,
VAR_X,
VAR_Y,
VAR_N,
#if FF_API_FRAME_PKT
VAR_POS,
#endif
VAR_T,
VAR_VARS_NB
};
2014-04-11 11:54:15 +02:00
typedef struct CropContext {
const AVClass *class;
int x; ///< x offset of the non-cropped area with respect to the input area
int y; ///< y offset of the non-cropped area with respect to the input area
int w; ///< width of the cropped area
int h; ///< height of the cropped area
2012-03-10 04:07:17 +01:00
AVRational out_sar; ///< output sample aspect ratio
int keep_aspect; ///< keep display aspect ratio when cropping
int exact; ///< exact cropping, for subsampled formats
2012-03-10 04:07:17 +01:00
int max_step[4]; ///< max pixel step for each plane, expressed as a number of bytes
int hsub, vsub; ///< chroma subsampling
char *x_expr, *y_expr, *w_expr, *h_expr;
AVExpr *x_pexpr, *y_pexpr; /* parsed expressions for x and y */
double var_values[VAR_VARS_NB];
} CropContext;
static int query_formats(AVFilterContext *ctx)
{
int reject_flags = AV_PIX_FMT_FLAG_BITSTREAM | FF_PIX_FMT_FLAG_SW_FLAT_SUB;
return ff_set_common_formats(ctx, ff_formats_pixdesc_filter(0, reject_flags));
}
static av_cold void uninit(AVFilterContext *ctx)
{
CropContext *s = ctx->priv;
av_expr_free(s->x_pexpr);
s->x_pexpr = NULL;
av_expr_free(s->y_pexpr);
s->y_pexpr = NULL;
}
static inline int normalize_double(int *n, double d)
{
int ret = 0;
if (isnan(d)) {
ret = AVERROR(EINVAL);
} else if (d > INT_MAX || d < INT_MIN) {
*n = d > INT_MAX ? INT_MAX : INT_MIN;
ret = AVERROR(EINVAL);
} else
*n = lrint(d);
return ret;
}
static int config_input(AVFilterLink *link)
{
AVFilterContext *ctx = link->dst;
CropContext *s = ctx->priv;
const AVPixFmtDescriptor *pix_desc = av_pix_fmt_desc_get(link->format);
int ret;
const char *expr;
double res;
s->var_values[VAR_IN_W] = s->var_values[VAR_IW] = ctx->inputs[0]->w;
s->var_values[VAR_IN_H] = s->var_values[VAR_IH] = ctx->inputs[0]->h;
s->var_values[VAR_A] = (float) link->w / link->h;
s->var_values[VAR_SAR] = link->sample_aspect_ratio.num ? av_q2d(link->sample_aspect_ratio) : 1;
s->var_values[VAR_DAR] = s->var_values[VAR_A] * s->var_values[VAR_SAR];
s->var_values[VAR_HSUB] = 1<<pix_desc->log2_chroma_w;
s->var_values[VAR_VSUB] = 1<<pix_desc->log2_chroma_h;
s->var_values[VAR_X] = NAN;
s->var_values[VAR_Y] = NAN;
s->var_values[VAR_OUT_W] = s->var_values[VAR_OW] = NAN;
s->var_values[VAR_OUT_H] = s->var_values[VAR_OH] = NAN;
s->var_values[VAR_N] = 0;
s->var_values[VAR_T] = NAN;
#if FF_API_FRAME_PKT
s->var_values[VAR_POS] = NAN;
#endif
av_image_fill_max_pixsteps(s->max_step, NULL, pix_desc);
if (pix_desc->flags & AV_PIX_FMT_FLAG_HWACCEL) {
s->hsub = 1;
s->vsub = 1;
} else {
s->hsub = pix_desc->log2_chroma_w;
s->vsub = pix_desc->log2_chroma_h;
}
av_expr_parse_and_eval(&res, (expr = s->w_expr),
var_names, s->var_values,
NULL, NULL, NULL, NULL, NULL, 0, ctx);
s->var_values[VAR_OUT_W] = s->var_values[VAR_OW] = res;
if ((ret = av_expr_parse_and_eval(&res, (expr = s->h_expr),
var_names, s->var_values,
2013-03-18 21:48:18 +01:00
NULL, NULL, NULL, NULL, NULL, 0, ctx)) < 0)
goto fail_expr;
s->var_values[VAR_OUT_H] = s->var_values[VAR_OH] = res;
/* evaluate again ow as it may depend on oh */
if ((ret = av_expr_parse_and_eval(&res, (expr = s->w_expr),
var_names, s->var_values,
2013-03-18 21:48:18 +01:00
NULL, NULL, NULL, NULL, NULL, 0, ctx)) < 0)
goto fail_expr;
s->var_values[VAR_OUT_W] = s->var_values[VAR_OW] = res;
if (normalize_double(&s->w, s->var_values[VAR_OUT_W]) < 0 ||
normalize_double(&s->h, s->var_values[VAR_OUT_H]) < 0) {
av_log(ctx, AV_LOG_ERROR,
"Too big value or invalid expression for out_w/ow or out_h/oh. "
"Maybe the expression for out_w:'%s' or for out_h:'%s' is self-referencing.\n",
s->w_expr, s->h_expr);
return AVERROR(EINVAL);
}
if (!s->exact) {
s->w &= ~((1 << s->hsub) - 1);
s->h &= ~((1 << s->vsub) - 1);
}
av_expr_free(s->x_pexpr);
av_expr_free(s->y_pexpr);
s->x_pexpr = s->y_pexpr = NULL;
if ((ret = av_expr_parse(&s->x_pexpr, s->x_expr, var_names,
NULL, NULL, NULL, NULL, 0, ctx)) < 0 ||
(ret = av_expr_parse(&s->y_pexpr, s->y_expr, var_names,
NULL, NULL, NULL, NULL, 0, ctx)) < 0)
return AVERROR(EINVAL);
if (s->keep_aspect) {
2012-03-10 04:07:17 +01:00
AVRational dar = av_mul_q(link->sample_aspect_ratio,
(AVRational){ link->w, link->h });
av_reduce(&s->out_sar.num, &s->out_sar.den,
dar.num * s->h, dar.den * s->w, INT_MAX);
2012-03-10 04:07:17 +01:00
} else
s->out_sar = link->sample_aspect_ratio;
2012-03-10 04:07:17 +01:00
av_log(ctx, AV_LOG_VERBOSE, "w:%d h:%d sar:%d/%d -> w:%d h:%d sar:%d/%d\n",
link->w, link->h, link->sample_aspect_ratio.num, link->sample_aspect_ratio.den,
s->w, s->h, s->out_sar.num, s->out_sar.den);
if (s->w <= 0 || s->h <= 0 ||
s->w > link->w || s->h > link->h) {
av_log(ctx, AV_LOG_ERROR,
"Invalid too big or non positive size for width '%d' or height '%d'\n",
s->w, s->h);
return AVERROR(EINVAL);
}
/* set default, required in the case the first computed value for x/y is NAN */
s->x = (link->w - s->w) / 2;
s->y = (link->h - s->h) / 2;
if (!s->exact) {
s->x &= ~((1 << s->hsub) - 1);
s->y &= ~((1 << s->vsub) - 1);
}
return 0;
fail_expr:
av_log(ctx, AV_LOG_ERROR, "Error when evaluating the expression '%s'\n", expr);
return ret;
}
static int config_output(AVFilterLink *link)
{
CropContext *s = link->src->priv;
const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(link->format);
if (desc->flags & AV_PIX_FMT_FLAG_HWACCEL) {
// Hardware frames adjust the cropping regions rather than
// changing the frame size.
} else {
link->w = s->w;
link->h = s->h;
}
link->sample_aspect_ratio = s->out_sar;
return 0;
}
static int filter_frame(AVFilterLink *link, AVFrame *frame)
{
AVFilterContext *ctx = link->dst;
CropContext *s = ctx->priv;
const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(link->format);
int i;
s->var_values[VAR_N] = link->frame_count_out;
s->var_values[VAR_T] = frame->pts == AV_NOPTS_VALUE ?
NAN : frame->pts * av_q2d(link->time_base);
#if FF_API_FRAME_PKT
FF_DISABLE_DEPRECATION_WARNINGS
s->var_values[VAR_POS] = frame->pkt_pos == -1 ?
NAN : frame->pkt_pos;
FF_ENABLE_DEPRECATION_WARNINGS
#endif
s->var_values[VAR_X] = av_expr_eval(s->x_pexpr, s->var_values, NULL);
s->var_values[VAR_Y] = av_expr_eval(s->y_pexpr, s->var_values, NULL);
/* It is necessary if x is expressed from y */
s->var_values[VAR_X] = av_expr_eval(s->x_pexpr, s->var_values, NULL);
normalize_double(&s->x, s->var_values[VAR_X]);
normalize_double(&s->y, s->var_values[VAR_Y]);
if (s->x < 0)
s->x = 0;
if (s->y < 0)
s->y = 0;
if ((unsigned)s->x + (unsigned)s->w > link->w)
s->x = link->w - s->w;
if ((unsigned)s->y + (unsigned)s->h > link->h)
s->y = link->h - s->h;
if (!s->exact) {
s->x &= ~((1 << s->hsub) - 1);
s->y &= ~((1 << s->vsub) - 1);
}
av_log(ctx, AV_LOG_TRACE, "n:%d t:%f x:%d y:%d x+w:%d y+h:%d\n",
(int)s->var_values[VAR_N], s->var_values[VAR_T],
s->x, s->y, s->x+s->w, s->y+s->h);
if (desc->flags & AV_PIX_FMT_FLAG_HWACCEL) {
frame->crop_top += s->y;
frame->crop_left += s->x;
frame->crop_bottom = frame->height - frame->crop_top - frame->crop_bottom - s->h;
frame->crop_right = frame->width - frame->crop_left - frame->crop_right - s->w;
} else {
frame->width = s->w;
frame->height = s->h;
frame->data[0] += s->y * frame->linesize[0];
frame->data[0] += s->x * s->max_step[0];
if (!(desc->flags & AV_PIX_FMT_FLAG_PAL)) {
for (i = 1; i < 3; i ++) {
if (frame->data[i]) {
frame->data[i] += (s->y >> s->vsub) * frame->linesize[i];
frame->data[i] += (s->x * s->max_step[i]) >> s->hsub;
}
}
}
/* alpha plane */
if (frame->data[3]) {
frame->data[3] += s->y * frame->linesize[3];
frame->data[3] += s->x * s->max_step[3];
}
}
return ff_filter_frame(link->dst->outputs[0], frame);
}
static int process_command(AVFilterContext *ctx, const char *cmd, const char *args,
char *res, int res_len, int flags)
{
CropContext *s = ctx->priv;
int ret;
if ( !strcmp(cmd, "out_w") || !strcmp(cmd, "w")
|| !strcmp(cmd, "out_h") || !strcmp(cmd, "h")
|| !strcmp(cmd, "x") || !strcmp(cmd, "y")) {
int old_x = s->x;
int old_y = s->y;
int old_w = s->w;
int old_h = s->h;
AVFilterLink *outlink = ctx->outputs[0];
AVFilterLink *inlink = ctx->inputs[0];
av_opt_set(s, cmd, args, 0);
if ((ret = config_input(inlink)) < 0) {
s->x = old_x;
s->y = old_y;
s->w = old_w;
s->h = old_h;
return ret;
}
ret = config_output(outlink);
} else
ret = AVERROR(ENOSYS);
return ret;
}
#define OFFSET(x) offsetof(CropContext, x)
#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
#define TFLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_RUNTIME_PARAM
static const AVOption crop_options[] = {
{ "out_w", "set the width crop area expression", OFFSET(w_expr), AV_OPT_TYPE_STRING, {.str = "iw"}, 0, 0, TFLAGS },
{ "w", "set the width crop area expression", OFFSET(w_expr), AV_OPT_TYPE_STRING, {.str = "iw"}, 0, 0, TFLAGS },
{ "out_h", "set the height crop area expression", OFFSET(h_expr), AV_OPT_TYPE_STRING, {.str = "ih"}, 0, 0, TFLAGS },
{ "h", "set the height crop area expression", OFFSET(h_expr), AV_OPT_TYPE_STRING, {.str = "ih"}, 0, 0, TFLAGS },
{ "x", "set the x crop area expression", OFFSET(x_expr), AV_OPT_TYPE_STRING, {.str = "(in_w-out_w)/2"}, 0, 0, TFLAGS },
{ "y", "set the y crop area expression", OFFSET(y_expr), AV_OPT_TYPE_STRING, {.str = "(in_h-out_h)/2"}, 0, 0, TFLAGS },
{ "keep_aspect", "keep aspect ratio", OFFSET(keep_aspect), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
{ "exact", "do exact cropping", OFFSET(exact), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
{ NULL }
};
AVFILTER_DEFINE_CLASS(crop);
static const AVFilterPad avfilter_vf_crop_inputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_VIDEO,
.filter_frame = filter_frame,
.config_props = config_input,
},
};
static const AVFilterPad avfilter_vf_crop_outputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_VIDEO,
.config_props = config_output,
},
};
const AVFilter ff_vf_crop = {
.name = "crop",
.description = NULL_IF_CONFIG_SMALL("Crop the input video."),
.priv_size = sizeof(CropContext),
.priv_class = &crop_class,
.uninit = uninit,
2021-08-12 13:05:31 +02:00
FILTER_INPUTS(avfilter_vf_crop_inputs),
FILTER_OUTPUTS(avfilter_vf_crop_outputs),
avfilter: Replace query_formats callback with union of list and callback If one looks at the many query_formats callbacks in existence, one will immediately recognize that there is one type of default callback for video and a slightly different default callback for audio: It is "return ff_set_common_formats_from_list(ctx, pix_fmts);" for video with a filter-specific pix_fmts list. For audio, it is the same with a filter-specific sample_fmts list together with ff_set_common_all_samplerates() and ff_set_common_all_channel_counts(). This commit allows to remove the boilerplate query_formats callbacks by replacing said callback with a union consisting the old callback and pointers for pixel and sample format arrays. For the not uncommon case in which these lists only contain a single entry (besides the sentinel) enum AVPixelFormat and enum AVSampleFormat fields are also added to the union to store them directly in the AVFilter, thereby avoiding a relocation. The state of said union will be contained in a new, dedicated AVFilter field (the nb_inputs and nb_outputs fields have been shrunk to uint8_t in order to create a hole for this new field; this is no problem, as the maximum of all the nb_inputs is four; for nb_outputs it is only two). The state's default value coincides with the earlier default of query_formats being unset, namely that the filter accepts all formats (and also sample rates and channel counts/layouts for audio) provided that these properties agree coincide for all inputs and outputs. By using different union members for audio and video filters the type-unsafety of using the same functions for audio and video lists will furthermore be more confined to formats.c than before. When the new fields are used, they will also avoid allocations: Currently something nearly equivalent to ff_default_query_formats() is called after every successful call to a query_formats callback; yet in the common case that the newly allocated AVFilterFormats are not used at all (namely if there are no free links) these newly allocated AVFilterFormats are freed again without ever being used. Filters no longer using the callback will not exhibit this any more. Reviewed-by: Paul B Mahol <onemda@gmail.com> Reviewed-by: Nicolas George <george@nsup.org> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2021-09-27 12:07:35 +02:00
FILTER_QUERY_FUNC(query_formats),
.process_command = process_command,
};