1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-28 20:53:54 +02:00
FFmpeg/libavfilter/vf_crop.c

392 lines
13 KiB
C
Raw 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
"pos", ///< position in the file
"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,
VAR_POS,
VAR_T,
VAR_VARS_NB
};
2014-04-11 12:54:15 +03:00
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 06:07:17 +03:00
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 06:07:17 +03:00
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)
{
AVFilterFormats *formats = NULL;
int fmt, ret;
for (fmt = 0; av_pix_fmt_desc_get(fmt); fmt++) {
const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(fmt);
if (!(desc->flags & (AV_PIX_FMT_FLAG_HWACCEL | AV_PIX_FMT_FLAG_BITSTREAM)) &&
!((desc->log2_chroma_w || desc->log2_chroma_h) && !(desc->flags & AV_PIX_FMT_FLAG_PLANAR)) &&
(ret = ff_add_format(&formats, fmt)) < 0)
return ret;
}
return ff_set_common_formats(ctx, formats);
}
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;
s->var_values[VAR_POS] = NAN;
av_image_fill_max_pixsteps(s->max_step, NULL, pix_desc);
s->hsub = pix_desc->log2_chroma_w;
s->vsub = pix_desc->log2_chroma_h;
if ((ret = av_expr_parse_and_eval(&res, (expr = s->w_expr),
var_names, s->var_values,
2013-03-18 23:48:18 +03:00
NULL, NULL, NULL, NULL, NULL, 0, ctx)) < 0)
goto fail_expr;
s->var_values[VAR_OUT_W] = s->var_values[VAR_OW] = res;
if ((ret = av_expr_parse_and_eval(&res, (expr = s->h_expr),
var_names, s->var_values,
2013-03-18 23:48:18 +03:00
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 23:48:18 +03:00
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 06:07:17 +03:00
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 06:07:17 +03:00
} else
s->out_sar = link->sample_aspect_ratio;
2012-03-10 06:07:17 +03:00
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(NULL, AV_LOG_ERROR, "Error when evaluating the expression '%s'\n", expr);
return ret;
}
static int config_output(AVFilterLink *link)
{
CropContext *s = link->src->priv;
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;
frame->width = s->w;
frame->height = s->h;
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);
s->var_values[VAR_POS] = frame->pkt_pos == -1 ?
NAN : frame->pkt_pos;
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 pos:%f x:%d y:%d x+w:%d y+h:%d\n",
(int)s->var_values[VAR_N], s->var_values[VAR_T], s->var_values[VAR_POS],
s->x, s->y, s->x+s->w, s->y+s->h);
frame->data[0] += s->y * frame->linesize[0];
frame->data[0] += s->x * s->max_step[0];
avutil/pixdesc: deprecate AV_PIX_FMT_FLAG_PSEUDOPAL PSEUDOPAL pixel formats are not paletted, but carried a palette with the intention of allowing code to treat unpaletted formats as paletted. The palette simply mapped the byte values to the resulting RGB values, making it some sort of LUT for RGB conversion. It was used for 1 byte formats only: RGB4_BYTE, BGR4_BYTE, RGB8, BGR8, GRAY8. The first 4 are awfully obscure, used only by some ancient bitmap formats. The last one, GRAY8, is more common, but its treatment is grossly incorrect. It considers full range GRAY8 only, so GRAY8 coming from typical Y video planes was not mapped to the correct RGB values. This cannot be fixed, because AVFrame.color_range can be freely changed at runtime, and there is nothing to ensure the pseudo palette is updated. Also, nothing actually used the PSEUDOPAL palette data, except xwdenc (trivially changed in the previous commit). All other code had to treat it as a special case, just to ignore or to propagate palette data. In conclusion, this was just a very strange old mechnaism that has no real justification to exist anymore (although it may have been nice and useful in the past). Now it's an artifact that makes the API harder to use: API users who allocate their own pixel data have to be aware that they need to allocate the palette, or FFmpeg will crash on them in _some_ situations. On top of this, there was no API to allocate the pseuo palette outside of av_frame_get_buffer(). This patch not only deprecates AV_PIX_FMT_FLAG_PSEUDOPAL, but also makes the pseudo palette optional. Nothing accesses it anymore, though if it's set, it's propagated. It's still allocated and initialized for compatibility with API users that rely on this feature. But new API users do not need to allocate it. This was an explicit goal of this patch. Most changes replace AV_PIX_FMT_FLAG_PSEUDOPAL with FF_PSEUDOPAL. I first tried #ifdefing all code, but it was a mess. The FF_PSEUDOPAL macro reduces the mess, and still allows defining FF_API_PSEUDOPAL to 0. Passes FATE with FF_API_PSEUDOPAL enabled and disabled. In addition, FATE passes with FF_API_PSEUDOPAL set to 1, but with allocation functions manually changed to not allocating a palette.
2018-03-29 15:18:28 +02:00
if (!(desc->flags & AV_PIX_FMT_FLAG_PAL || desc->flags & FF_PSEUDOPAL)) {
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
static const AVOption crop_options[] = {
{ "out_w", "set the width crop area expression", OFFSET(w_expr), AV_OPT_TYPE_STRING, {.str = "iw"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "w", "set the width crop area expression", OFFSET(w_expr), AV_OPT_TYPE_STRING, {.str = "iw"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "out_h", "set the height crop area expression", OFFSET(h_expr), AV_OPT_TYPE_STRING, {.str = "ih"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "h", "set the height crop area expression", OFFSET(h_expr), AV_OPT_TYPE_STRING, {.str = "ih"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "x", "set the x crop area expression", OFFSET(x_expr), AV_OPT_TYPE_STRING, {.str = "(in_w-out_w)/2"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "y", "set the y crop area expression", OFFSET(y_expr), AV_OPT_TYPE_STRING, {.str = "(in_h-out_h)/2"}, CHAR_MIN, CHAR_MAX, FLAGS },
{ "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,
},
{ NULL }
};
static const AVFilterPad avfilter_vf_crop_outputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_VIDEO,
.config_props = config_output,
},
{ NULL }
};
AVFilter ff_vf_crop = {
.name = "crop",
.description = NULL_IF_CONFIG_SMALL("Crop the input video."),
.priv_size = sizeof(CropContext),
.priv_class = &crop_class,
.query_formats = query_formats,
.uninit = uninit,
.inputs = avfilter_vf_crop_inputs,
.outputs = avfilter_vf_crop_outputs,
.process_command = process_command,
};