lavfi: add video stabilization plugins using vid.stab library

vidstabdetect and vidstabtransform common functions for interfacing
vid.stab are in libavfilter/vidstabutils.c

Signed-off-by: Georg Martius <martius@mis.mpg.de>

Changelog

@@ -31,6 +31,8 @@ version <next>:
 - asetrate filter
 - interleave filter
 - timeline editing with filters
+- vidstabdetect and vidstabtransform filters for video stabilization using
+  the vid.stab library
 
 
 version 1.2:

LICENSE

@@ -72,6 +72,7 @@ compatible libraries
 The following libraries are under GPL:
     - libcdio
     - libutvideo
+    - libvidstab
     - libx264
     - libxavs
     - libxvid

configure

@@ -226,6 +226,7 @@ External library support:
   --enable-libtwolame      enable MP2 encoding via libtwolame [no]
   --enable-libutvideo      enable Ut Video encoding and decoding via libutvideo [no]
   --enable-libv4l2         enable libv4l2/v4l-utils [no]
+  --enable-libvidstab      enable video stabilization using vid.stab [no]
   --enable-libvo-aacenc    enable AAC encoding via libvo-aacenc [no]
   --enable-libvo-amrwbenc  enable AMR-WB encoding via libvo-amrwbenc [no]
   --enable-libvorbis       enable Vorbis en/decoding via libvorbis,
@@ -1181,6 +1182,7 @@ EXTERNAL_LIBRARY_LIST="
     libtwolame
     libutvideo
     libv4l2
+    libvidstab
     libvo_aacenc
     libvo_amrwbenc
     libvorbis
@@ -2152,6 +2154,8 @@ stereo3d_filter_deps="gpl"
 subtitles_filter_deps="avformat avcodec libass"
 super2xsai_filter_deps="gpl"
 tinterlace_filter_deps="gpl"
+vidstabdetect_filter_deps="libvidstab"
+vidstabtransform_filter_deps="libvidstab"
 yadif_filter_deps="gpl"
 pixfmts_super2xsai_test_deps="super2xsai_filter"
 tinterlace_merge_test_deps="tinterlace_filter"
@@ -3573,6 +3577,7 @@ die_license_disabled_gpl() {
 
 die_license_disabled gpl libcdio
 die_license_disabled gpl libutvideo
+die_license_disabled gpl libvidstab
 die_license_disabled gpl libx264
 die_license_disabled gpl libxavs
 die_license_disabled gpl libxvid
@@ -4015,6 +4020,7 @@ enabled libtwolame && require  libtwolame twolame.h twolame_init -ltwolame &&
                         die "ERROR: libtwolame must be installed and version must be >= 0.3.10"; }
 enabled libutvideo    && require_cpp utvideo "stdint.h stdlib.h utvideo/utvideo.h utvideo/Codec.h" 'CCodec*' -lutvideo -lstdc++
 enabled libv4l2    && require_pkg_config libv4l2 libv4l2.h v4l2_ioctl
+enabled libvidstab && require_pkg_config vidstab vid.stab/libvidstab.h vsMotionDetectInit
 enabled libvo_aacenc && require libvo_aacenc vo-aacenc/voAAC.h voGetAACEncAPI -lvo-aacenc
 enabled libvo_amrwbenc && require libvo_amrwbenc vo-amrwbenc/enc_if.h E_IF_init -lvo-amrwbenc
 enabled libvorbis  && require  libvorbis vorbis/vorbisenc.h vorbis_info_init -lvorbisenc -lvorbis -logg

doc/filters.texi

@@ -5308,6 +5308,141 @@ much, but it will increase the amount of blurring needed to cover over
 the image and will destroy more information than necessary, and extra
 pixels will slow things down on a large logo.
 
+@anchor{vidstabtransform}
+@section vidstabtransform
+
+Video stabilization/deshaking: pass 2 of 2,
+see @ref{vidstabdetect} for pass 1.
+
+Read a file with transform information for each frame and
+apply/compensate them. Together with the @ref{vidstabdetect}
+filter this can be used to deshake videos. See also
+@url{http://public.hronopik.de/vid.stab}. It is important to also use
+the unsharp filter, see below.
+
+To enable compilation of this filter you need to configure FFmpeg with
+@code{--enable-libvidstab}.
+
+This filter accepts the following named options, expressed as a
+sequence of @var{key}=@var{value} pairs, separated by ":".
+
+@table @option
+
+@item input
+path to the file used to read the transforms (default: @file{transforms.trf})
+
+@item smoothing
+number of frames (value*2 + 1) used for lowpass filtering the camera movements
+(default: 10). For example a number of 10 means that 21 frames are used
+(10 in the past and 10 in the future) to smoothen the motion in the
+video. A larger values leads to a smoother video, but limits the
+acceleration of the camera (pan/tilt movements).
+
+@item maxshift
+maximal number of pixels to translate frames (default: -1 no limit)
+
+@item maxangle
+maximal angle in radians (degree*PI/180) to rotate frames (default: -1
+no limit)
+
+@item crop
+How to deal with borders that may be visible due to movement
+compensation. Available values are:
+
+@table @samp
+@item keep
+keep image information from previous frame (default)
+@item black
+fill the border black
+@end table
+
+@item invert
+@table @samp
+@item 0
+ keep transforms normal (default)
+@item 1
+ invert transforms
+@end table
+
+
+@item relative
+consider transforms as
+@table @samp
+@item 0
+ absolute
+@item 1
+ relative to previous frame (default)
+@end table
+
+
+@item zoom
+percentage to zoom (default: 0)
+@table @samp
+@item >0
+  zoom in
+@item <0
+  zoom out
+@end table
+
+@item optzoom
+if 1 then optimal zoom value is determined (default).
+Optimal zoom means no (or only little) border should be visible.
+Note that the value given at zoom is added to the one calculated
+here.
+
+@item interpol
+type of interpolation
+
+Available values are:
+@table @samp
+@item no
+no interpolation
+@item linear
+linear only horizontal
+@item bilinear
+linear in both directions (default)
+@item bicubic
+cubic in both directions (slow)
+@end table
+
+@item tripod
+virtual tripod mode means that the video is stabilized such that the
+camera stays stationary. Use also @code{tripod} option of
+@ref{vidstabdetect}.
+@table @samp
+@item 0
+off (default)
+@item 1
+virtual tripod mode: equivalent to @code{relative=0:smoothing=0}
+@end table
+
+@end table
+
+@subsection Examples
+
+@itemize
+@item
+typical call with default default values:
+ (note the unsharp filter which is always recommended)
+@example
+ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
+@end example
+
+@item
+zoom in a bit more and load transform data from a given file
+@example
+vidstabtransform=zoom=5:input="mytransforms.trf"
+@end example
+
+@item
+smoothen the video even more
+@example
+vidstabtransform=smoothing=30
+@end example
+
+@end itemize
+
+
 @section scale
 
 Scale (resize) the input video, using the libswscale library.
@@ -5706,6 +5841,93 @@ in [-30,0] will filter edges. Default value is 0.
 If a chroma option is not explicitly set, the corresponding luma value
 is set.
 
+@anchor{vidstabdetect}
+@section vidstabdetect
+
+Video stabilization/deshaking: pass 1 of 2, see @ref{vidstabtransform}
+for pass 2.
+
+Generates a file with relative transform information translation,
+rotation about subsequent frames.
+
+To enable compilation of this filter you need to configure FFmpeg with
+@code{--enable-libvidstab}.
+
+This filter accepts the following named options, expressed as a
+sequence of @var{key}=@var{value} pairs, separated by ":".
+
+@table @option
+@item result
+path to the file used to write the transforms (default:@file{transforms.trf})
+
+@item shakiness
+how shaky is the video and how quick is the camera? (default: 5)
+@table @samp
+@item 1
+ little (fast)
+@item ...
+@item 10
+ very strong/quick (slow)
+@end table
+
+@item accuracy
+accuracy of detection process (>=shakiness) (default: 9)
+@table @samp
+@item 1
+ low (fast)
+@item 15
+ high (slow)
+@end table
+
+@item stepsize
+stepsize of search process, region around minimum is scanned with 1 pixel
+resolution (default: 6)
+
+@item mincontrast
+below this contrast a local measurement field is discarded (0-1) (default: 0.3)
+
+@item tripod
+virtual tripod mode: @code{tripod=framenum} if framenum>0 otherwise disabled.
+The motion of the frames is compared to a reference frame (framenum).
+The idea is to compensate all movements in a more-or-less static scene
+ and keep the camera view absolutely still.
+(default: 0 (disabled))
+
+@item show
+draw nothing (default); 1,2: show fields and transforms in the resulting frames
+
+@end table
+
+@subsection Examples
+
+@itemize
+@item
+use default values:
+@example
+vidstabdetect
+@end example
+
+@item
+strongly shaky movie and put the results in @code{mytransforms.trf}
+@example
+vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
+@end example
+
+@item
+visualize some internals in the resulting video
+@example
+vidstabdetect=show=1
+@end example
+
+
+@item
+Typical call with visualization
+@example
+ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
+@end example
+@end itemize
+
+
 @section stereo3d
 
 Convert between different stereoscopic image formats.

libavfilter/Makefile

@@ -179,6 +179,8 @@ OBJS-$(CONFIG_TINTERLACE_FILTER)             += vf_tinterlace.o
 OBJS-$(CONFIG_TRANSPOSE_FILTER)              += vf_transpose.o
 OBJS-$(CONFIG_UNSHARP_FILTER)                += vf_unsharp.o
 OBJS-$(CONFIG_VFLIP_FILTER)                  += vf_vflip.o
+OBJS-$(CONFIG_VIDSTABDETECT_FILTER)          += vidstabutils.o vf_vidstabdetect.o
+OBJS-$(CONFIG_VIDSTABTRANSFORM_FILTER)       += vidstabutils.o vf_vidstabtransform.o
 OBJS-$(CONFIG_YADIF_FILTER)                  += vf_yadif.o
 
 OBJS-$(CONFIG_CELLAUTO_FILTER)               += vsrc_cellauto.o

libavfilter/allfilters.c

@@ -176,6 +176,8 @@ void avfilter_register_all(void)
     REGISTER_FILTER(TRANSPOSE,      transpose,      vf);
     REGISTER_FILTER(UNSHARP,        unsharp,        vf);
     REGISTER_FILTER(VFLIP,          vflip,          vf);
+    REGISTER_FILTER(VIDSTABDETECT,  vidstabdetect,  vf);
+    REGISTER_FILTER(VIDSTABTRANSFORM, vidstabtransform, vf);
     REGISTER_FILTER(YADIF,          yadif,          vf);
 
     REGISTER_FILTER(CELLAUTO,       cellauto,       vsrc);

libavfilter/vf_vidstabdetect.c

@@ -0,0 +1,239 @@
+/*
+ * Copyright (c) 2013 Georg Martius <georg dot martius at web dot de>
+ *
+ * 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
+ */
+
+#define DEFAULT_RESULT_NAME     "transforms.trf"
+
+#include <vid.stab/libvidstab.h>
+
+#include "libavutil/common.h"
+#include "libavutil/opt.h"
+#include "libavutil/imgutils.h"
+#include "avfilter.h"
+#include "internal.h"
+
+#include "vidstabutils.h"
+
+typedef struct {
+    const AVClass* class;
+
+    VSMotionDetect md;
+    VSMotionDetectConfig conf;
+
+    char* result;
+    FILE* f;
+} StabData;
+
+
+#define OFFSET(x) offsetof(StabData, x)
+#define OFFSETC(x) (offsetof(StabData, conf)+offsetof(VSMotionDetectConfig, x))
+#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
+
+static const AVOption vidstabdetect_options[]= {
+    {"result",      "path to the file used to write the transforms (def:transforms.trf)", OFFSET(result),              AV_OPT_TYPE_STRING, {.str = DEFAULT_RESULT_NAME}},
+    {"shakiness",   "how shaky is the video and how quick is the camera?"
+                    " 1: little (fast) 10: very strong/quick (slow) (def: 5)",            OFFSETC(shakiness),         AV_OPT_TYPE_INT,    {.i64 = 5},  1, 10,       FLAGS},
+    {"accuracy",    "(>=shakiness) 1: low 15: high (slow) (def: 9)",                      OFFSETC(accuracy),          AV_OPT_TYPE_INT,    {.i64 = 9 }, 1, 15,       FLAGS},
+    {"stepsize",    "region around minimum is scanned with 1 pixel resolution (def: 6)",  OFFSETC(stepSize),          AV_OPT_TYPE_INT,    {.i64 = 6},  1, 32,       FLAGS},
+    {"mincontrast", "below this contrast a field is discarded (0-1) (def: 0.3)",          OFFSETC(contrastThreshold), AV_OPT_TYPE_DOUBLE, {.dbl =  0.25}, 0.0, 1.0, FLAGS},
+    {"show",        "0: draw nothing (def); 1,2: show fields and transforms",             OFFSETC(show),              AV_OPT_TYPE_INT,    {.i64 =  0}, 0, 2,        FLAGS},
+    {"tripod",      "virtual tripod mode (if >0): motion is compared to a reference"
+                    " reference frame (frame # is the value) (def: 0)",                   OFFSETC(virtualTripod),     AV_OPT_TYPE_INT,    {.i64 = 0},  0, INT_MAX,  FLAGS},
+    {NULL},
+};
+
+AVFILTER_DEFINE_CLASS(vidstabdetect);
+
+static av_cold int init(AVFilterContext *ctx)
+{
+    StabData* sd = ctx->priv;
+    vs_set_mem_and_log_functions();
+    sd->class = &vidstabdetect_class;
+    av_log(ctx, AV_LOG_VERBOSE, "vidstabdetect filter: init %s\n", LIBVIDSTAB_VERSION);
+    return 0;
+}
+
+static av_cold void uninit(AVFilterContext *ctx)
+{
+    StabData *sd = ctx->priv;
+    VSMotionDetect* md = &(sd->md);
+
+    if (sd->f) {
+        fclose(sd->f);
+        sd->f = NULL;
+    }
+
+    vsMotionDetectionCleanup(md);
+
+}
+
+static int query_formats(AVFilterContext *ctx)
+{
+    // If you add something here also add it in vidstabutils.c
+    static const enum AVPixelFormat pix_fmts[] = {
+        AV_PIX_FMT_YUV444P,  AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUV420P,
+        AV_PIX_FMT_YUV411P,  AV_PIX_FMT_YUV410P, AV_PIX_FMT_YUVA420P,
+        AV_PIX_FMT_YUV440P,  AV_PIX_FMT_GRAY8,
+        AV_PIX_FMT_RGB24, AV_PIX_FMT_BGR24, AV_PIX_FMT_RGBA,
+        AV_PIX_FMT_NONE
+    };
+
+    ff_set_common_formats(ctx, ff_make_format_list(pix_fmts));
+    return 0;
+}
+
+
+static int config_input(AVFilterLink *inlink)
+{
+    AVFilterContext *ctx = inlink->dst;
+    StabData *sd = ctx->priv;
+
+    VSMotionDetect* md = &(sd->md);
+    VSFrameInfo fi;
+    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
+
+    vsFrameInfoInit(&fi,inlink->w, inlink->h, av_2_vs_pixel_format(ctx, inlink->format));
+    if(fi.bytesPerPixel != av_get_bits_per_pixel(desc)/8){
+        av_log(ctx, AV_LOG_ERROR, "pixel-format error: wrong bits/per/pixel, please report a BUG");
+        return AVERROR(EINVAL);
+    }
+    if(fi.log2ChromaW != desc->log2_chroma_w){
+        av_log(ctx, AV_LOG_ERROR, "pixel-format error: log2_chroma_w, please report a BUG");
+        return AVERROR(EINVAL);
+    }
+
+    if(fi.log2ChromaH != desc->log2_chroma_h){
+        av_log(ctx, AV_LOG_ERROR, "pixel-format error: log2_chroma_h, please report a BUG");
+        return AVERROR(EINVAL);
+    }
+
+    // set values that are not initializes by the options
+    sd->conf.algo     = 1;
+    sd->conf.modName  = "vidstabdetect";
+    if(vsMotionDetectInit(md, &sd->conf, &fi) != VS_OK){
+        av_log(ctx, AV_LOG_ERROR, "initialization of Motion Detection failed, please report a BUG");
+        return AVERROR(EINVAL);
+    }
+
+    vsMotionDetectGetConfig(&sd->conf, md);
+    av_log(ctx, AV_LOG_INFO, "Video stabilization settings (pass 1/2):\n");
+    av_log(ctx, AV_LOG_INFO, "     shakiness = %d\n", sd->conf.shakiness);
+    av_log(ctx, AV_LOG_INFO, "      accuracy = %d\n", sd->conf.accuracy);
+    av_log(ctx, AV_LOG_INFO, "      stepsize = %d\n", sd->conf.stepSize);
+    av_log(ctx, AV_LOG_INFO, "   mincontrast = %f\n", sd->conf.contrastThreshold);
+    av_log(ctx, AV_LOG_INFO, "          show = %d\n", sd->conf.show);
+    av_log(ctx, AV_LOG_INFO, "        result = %s\n", sd->result);
+
+    sd->f = fopen(sd->result, "w");
+    if (sd->f == NULL) {
+        av_log(ctx, AV_LOG_ERROR, "cannot open transform file %s\n", sd->result);
+        return AVERROR(EINVAL);
+    }else{
+        if(vsPrepareFile(md, sd->f) != VS_OK){
+            av_log(ctx, AV_LOG_ERROR, "cannot write to transform file %s\n", sd->result);
+            return AVERROR(EINVAL);
+        }
+    }
+    return 0;
+}
+
+
+static int filter_frame(AVFilterLink *inlink, AVFrame *in)
+{
+    AVFilterContext *ctx = inlink->dst;
+    StabData *sd = ctx->priv;
+    VSMotionDetect* md = &(sd->md);
+    LocalMotions localmotions;
+
+    AVFilterLink *outlink = inlink->dst->outputs[0];
+    int direct = 0;
+    AVFrame *out;
+    VSFrame frame;
+    int plane;
+
+    if (av_frame_is_writable(in)) {
+        direct = 1;
+        out = in;
+    } else {
+        out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
+        if (!out) {
+            av_frame_free(&in);
+            return AVERROR(ENOMEM);
+        }
+        av_frame_copy_props(out, in);
+    }
+
+    for(plane=0; plane < md->fi.planes; plane++){
+        frame.data[plane] = in->data[plane];
+        frame.linesize[plane] = in->linesize[plane];
+    }
+    if(vsMotionDetection(md, &localmotions, &frame) !=  VS_OK){
+        av_log(ctx, AV_LOG_ERROR, "motion detection failed");
+        return AVERROR(AVERROR_EXTERNAL);
+    } else {
+        if(vsWriteToFile(md, sd->f, &localmotions) != VS_OK){
+            av_log(ctx, AV_LOG_ERROR, "cannot write to transform file");
+            return AVERROR(errno);
+        }
+        vs_vector_del(&localmotions);
+    }
+    if(sd->conf.show>0 && !direct){
+        av_image_copy(out->data, out->linesize,
+                      (void*)in->data, in->linesize,
+                      in->format, in->width, in->height);
+    }
+
+    if (!direct)
+        av_frame_free(&in);
+
+    return ff_filter_frame(outlink, out);
+}
+
+static const AVFilterPad avfilter_vf_vidstabdetect_inputs[] = {
+    {
+        .name             = "default",
+        .type             = AVMEDIA_TYPE_VIDEO,
+        .filter_frame     = filter_frame,
+        .config_props     = config_input,
+    },
+    { NULL }
+};
+
+static const AVFilterPad avfilter_vf_vidstabdetect_outputs[] = {
+    {
+        .name             = "default",
+        .type             = AVMEDIA_TYPE_VIDEO,
+    },
+    { NULL }
+};
+
+AVFilter avfilter_vf_vidstabdetect = {
+    .name          = "vidstabdetect",
+    .description   = NULL_IF_CONFIG_SMALL("pass 1 of 2 for stabilization"
+                                          "extracts relative transformations"
+                                          "(pass 2 see vidstabtransform)"),
+    .priv_size     = sizeof(StabData),
+    .init          = init,
+    .uninit        = uninit,
+    .query_formats = query_formats,
+
+    .inputs        = avfilter_vf_vidstabdetect_inputs,
+    .outputs       = avfilter_vf_vidstabdetect_outputs,
+    .priv_class    = &vidstabdetect_class,
+};

libavfilter/vf_vidstabtransform.c

@@ -0,0 +1,294 @@
+/*
+ * Copyright (c) 2013 Georg Martius <georg dot martius at web dot de>
+ *
+ * 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
+ */
+
+#define DEFAULT_INPUT_NAME     "transforms.trf"
+
+#include <vid.stab/libvidstab.h>
+
+#include "libavutil/common.h"
+#include "libavutil/opt.h"
+#include "libavutil/imgutils.h"
+#include "avfilter.h"
+#include "internal.h"
+
+#include "vidstabutils.h"
+
+typedef struct {
+    const AVClass* class;
+
+    VSTransformData td;
+    VSTransformConfig conf;
+
+    VSTransformations trans; // transformations
+    char* input;           // name of transform file
+    int tripod;
+} TransformContext;
+
+#define OFFSET(x) offsetof(TransformContext, x)
+#define OFFSETC(x) (offsetof(TransformContext, conf)+offsetof(VSTransformConfig, x))
+#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
+
+static const AVOption vidstabtransform_options[]= {
+    {"input",     "path to the file storing the transforms (def:transforms.trf)",   OFFSET(input),
+                   AV_OPT_TYPE_STRING, {.str = DEFAULT_INPUT_NAME} },
+    {"smoothing", "number of frames*2 + 1 used for lowpass filtering (def: 10)",    OFFSETC(smoothing),
+                   AV_OPT_TYPE_INT,    {.i64 = 10},       1, 1000, FLAGS},
+    {"maxshift",  "maximal number of pixels to translate image (def: -1 no limit)", OFFSETC(maxShift),
+                   AV_OPT_TYPE_INT,    {.i64 = -1},      -1, 500,  FLAGS},
+    {"maxangle",  "maximal angle in rad to rotate image (def: -1 no limit)",        OFFSETC(maxAngle),
+                   AV_OPT_TYPE_DOUBLE, {.dbl = -1.0},  -1.0, 3.14, FLAGS},
+    {"crop",      "keep: (def), black",                                             OFFSETC(crop),
+                   AV_OPT_TYPE_INT,    {.i64 = 0},        0, 1,    FLAGS, "crop"},
+    {  "keep",    "keep border",                                                    0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VSKeepBorder }, 0, 0, FLAGS, "crop"},
+    {  "black",   "black border",                                                   0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VSCropBorder }, 0, 0, FLAGS, "crop"},
+    {"invert",    "1: invert transforms (def: 0)",                                  OFFSETC(invert),
+                   AV_OPT_TYPE_INT,    {.i64 = 0},        0, 1,    FLAGS},
+    {"relative",  "consider transforms as 0: abslute, 1: relative (def)",          OFFSETC(relative),
+                   AV_OPT_TYPE_INT,    {.i64 = 1},        0, 1,    FLAGS},
+    {"zoom",      "percentage to zoom >0: zoom in, <0 zoom out (def: 0)",           OFFSETC(zoom),
+                   AV_OPT_TYPE_DOUBLE, {.dbl = 0},        0, 100,  FLAGS},
+    {"optzoom",   "0: nothing, 1: determine optimal zoom (def) (added to 'zoom')",  OFFSETC(optZoom),
+                   AV_OPT_TYPE_INT,    {.i64 = 1},        0, 1,    FLAGS},
+    {"interpol",  "type of interpolation, no, linear, bilinear (def) , bicubic",    OFFSETC(interpolType),
+                   AV_OPT_TYPE_INT,    {.i64 = 2},        0, 3,    FLAGS, "interpol"},
+    {  "no",      "no interpolation",                                               0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VS_Zero  },  0, 0,  FLAGS, "interpol"},
+    {  "linear",  "linear (horizontal)",                                            0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VS_Linear }, 0, 0,  FLAGS, "interpol"},
+    {  "bilinear","bi-linear",                                                      0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VS_BiLinear},0, 0,  FLAGS, "interpol"},
+    {  "bicubic", "bi-cubic",                                                       0,
+                   AV_OPT_TYPE_CONST,  {.i64 = VS_BiCubic },0, 0,  FLAGS, "interpol"},
+    {"tripod",    "if 1: virtual tripod mode (equiv. to relative=0:smoothing=0)",   OFFSET(tripod),
+                   AV_OPT_TYPE_INT,    {.i64 = 0},        0, 1,    FLAGS},
+    {NULL},
+};
+
+AVFILTER_DEFINE_CLASS(vidstabtransform);
+
+static av_cold int init(AVFilterContext *ctx)
+{
+    TransformContext* tc = ctx->priv;
+    vs_set_mem_and_log_functions();
+    tc->class = &vidstabtransform_class;
+    av_log(ctx, AV_LOG_VERBOSE, "vidstabtransform filter: init %s\n", LIBVIDSTAB_VERSION);
+    return 0;
+}
+
+static av_cold void uninit(AVFilterContext *ctx)
+{
+    TransformContext *tc = ctx->priv;
+
+    vsTransformDataCleanup(&tc->td);
+    vsTransformationsCleanup(&tc->trans);
+}
+
+static int query_formats(AVFilterContext *ctx)
+{
+    // If you add something here also add it in vidstabutils.c
+    static const enum AVPixelFormat pix_fmts[] = {
+        AV_PIX_FMT_YUV444P,  AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUV420P,
+        AV_PIX_FMT_YUV411P,  AV_PIX_FMT_YUV410P, AV_PIX_FMT_YUVA420P,
+        AV_PIX_FMT_YUV440P,  AV_PIX_FMT_GRAY8,
+        AV_PIX_FMT_RGB24, AV_PIX_FMT_BGR24, AV_PIX_FMT_RGBA,
+        AV_PIX_FMT_NONE
+    };
+
+    ff_set_common_formats(ctx, ff_make_format_list(pix_fmts));
+    return 0;
+}
+
+
+static int config_input(AVFilterLink *inlink)
+{
+    AVFilterContext *ctx = inlink->dst;
+    TransformContext *tc = ctx->priv;
+    FILE* f;
+
+    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
+
+    VSTransformData* td = &(tc->td);
+
+    VSFrameInfo fi_src;
+    VSFrameInfo fi_dest;
+
+    if(!vsFrameInfoInit(&fi_src, inlink->w, inlink->h,
+                      av_2_vs_pixel_format(ctx,inlink->format)) ||
+       !vsFrameInfoInit(&fi_dest, inlink->w, inlink->h,
+                      av_2_vs_pixel_format(ctx, inlink->format))){
+        av_log(ctx, AV_LOG_ERROR, "unknown pixel format: %i (%s)",
+               inlink->format, desc->name);
+        return AVERROR(EINVAL);
+    }
+
+    if(fi_src.bytesPerPixel != av_get_bits_per_pixel(desc)/8 ||
+       fi_src.log2ChromaW != desc->log2_chroma_w ||
+       fi_src.log2ChromaH != desc->log2_chroma_h){
+        av_log(ctx, AV_LOG_ERROR, "pixel-format error: bpp %i<>%i  ",
+               fi_src.bytesPerPixel, av_get_bits_per_pixel(desc)/8);
+        av_log(ctx, AV_LOG_ERROR, "chroma_subsampl: w: %i<>%i  h: %i<>%i\n",
+               fi_src.log2ChromaW, desc->log2_chroma_w,
+               fi_src.log2ChromaH, desc->log2_chroma_h);
+        return AVERROR(EINVAL);
+    }
+
+    // set values that are not initializes by the options
+    tc->conf.modName = "vidstabtransform";
+    tc->conf.verbose =1;
+    if(tc->tripod){
+        av_log(ctx, AV_LOG_INFO, "Virtual tripod mode: relative=0, smoothing=0");
+        tc->conf.relative=0;
+        tc->conf.smoothing=0;
+    }
+
+    if(vsTransformDataInit(td, &tc->conf, &fi_src, &fi_dest) != VS_OK){
+        av_log(ctx, AV_LOG_ERROR, "initialization of vid.stab transform failed, please report a BUG\n");
+        return AVERROR(EINVAL);
+    }
+
+    vsTransformGetConfig(&tc->conf,td);
+    av_log(ctx, AV_LOG_INFO, "Video transformation/stabilization settings (pass 2/2):\n");
+    av_log(ctx, AV_LOG_INFO, "    input     = %s\n", tc->input);
+    av_log(ctx, AV_LOG_INFO, "    smoothing = %d\n", tc->conf.smoothing);
+    av_log(ctx, AV_LOG_INFO, "    maxshift  = %d\n", tc->conf.maxShift);
+    av_log(ctx, AV_LOG_INFO, "    maxangle  = %f\n", tc->conf.maxAngle);
+    av_log(ctx, AV_LOG_INFO, "    crop      = %s\n", tc->conf.crop ? "Black" : "Keep");
+    av_log(ctx, AV_LOG_INFO, "    relative  = %s\n", tc->conf.relative ? "True": "False");
+    av_log(ctx, AV_LOG_INFO, "    invert    = %s\n", tc->conf.invert ? "True" : "False");
+    av_log(ctx, AV_LOG_INFO, "    zoom      = %f\n", tc->conf.zoom);
+    av_log(ctx, AV_LOG_INFO, "    optzoom   = %s\n", tc->conf.optZoom ? "On" : "Off");
+    av_log(ctx, AV_LOG_INFO, "    interpol  = %s\n", getInterpolationTypeName(tc->conf.interpolType));
+
+    f = fopen(tc->input, "r");
+    if (f == NULL) {
+        av_log(ctx, AV_LOG_ERROR, "cannot open input file %s\n", tc->input);
+        return AVERROR(errno);
+    } else {
+        VSManyLocalMotions mlms;
+        if(vsReadLocalMotionsFile(f,&mlms)==VS_OK){
+            // calculate the actual transforms from the localmotions
+            if(vsLocalmotions2TransformsSimple(td, &mlms,&tc->trans)!=VS_OK){
+                av_log(ctx, AV_LOG_ERROR, "calculating transformations failed\n");
+                return AVERROR(EINVAL);
+            }
+        }else{ // try to read old format
+            if (!vsReadOldTransforms(td, f, &tc->trans)) { /* read input file */
+                av_log(ctx, AV_LOG_ERROR, "error parsing input file %s\n", tc->input);
+                return AVERROR(EINVAL);
+            }
+        }
+    }
+    fclose(f);
+
+    if (vsPreprocessTransforms(td, &tc->trans)!= VS_OK ) {
+        av_log(ctx, AV_LOG_ERROR, "error while preprocessing transforms\n");
+        return AVERROR(EINVAL);
+    }
+
+    // TODO: add sharpening, so far the user needs to call the unsharp filter manually
+    return 0;
+}
+
+
+static int filter_frame(AVFilterLink *inlink,  AVFrame *in)
+{
+    AVFilterContext *ctx = inlink->dst;
+    TransformContext *tc = ctx->priv;
+    VSTransformData* td = &(tc->td);
+
+    AVFilterLink *outlink = inlink->dst->outputs[0];
+    int direct = 0;
+    AVFrame *out;
+    VSFrame inframe;
+    int plane;
+
+    if (av_frame_is_writable(in)) {
+        direct = 1;
+        out = in;
+    } else {
+        out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
+        if (!out) {
+            av_frame_free(&in);
+            return AVERROR(ENOMEM);
+        }
+        av_frame_copy_props(out, in);
+    }
+
+    for(plane=0; plane < vsTransformGetSrcFrameInfo(td)->planes; plane++){
+        inframe.data[plane] = in->data[plane];
+        inframe.linesize[plane] = in->linesize[plane];
+    }
+    if(out == in){ // inplace
+        vsTransformPrepare(td, &inframe, &inframe);
+    }else{ // seperate frames
+        VSFrame outframe;
+        for(plane=0; plane < vsTransformGetDestFrameInfo(td)->planes; plane++){
+            outframe.data[plane] = out->data[plane];
+            outframe.linesize[plane] = out->linesize[plane];
+        }
+        vsTransformPrepare(td, &inframe, &outframe);
+    }
+
+    vsDoTransform(td, vsGetNextTransform(td, &tc->trans));
+
+    vsTransformFinish(td);
+
+    if (!direct)
+        av_frame_free(&in);
+
+    return ff_filter_frame(outlink, out);
+}
+
+static const AVFilterPad avfilter_vf_vidstabtransform_inputs[] = {
+    {
+        .name             = "default",
+        .type             = AVMEDIA_TYPE_VIDEO,
+        .filter_frame     = filter_frame,
+        .config_props     = config_input,
+    },
+    { NULL }
+};
+
+static const AVFilterPad avfilter_vf_vidstabtransform_outputs[] = {
+    {
+        .name             = "default",
+        .type             = AVMEDIA_TYPE_VIDEO,
+    },
+    { NULL }
+};
+
+AVFilter avfilter_vf_vidstabtransform = {
+    .name          = "vidstabtransform",
+    .description   = NULL_IF_CONFIG_SMALL("pass 2 of stabilization"
+                                          "transforms the frames"
+                                          "(see vidstabdetect for pass 1)"),
+    .priv_size     = sizeof(TransformContext),
+    .init          = init,
+    .uninit        = uninit,
+    .query_formats = query_formats,
+
+    .inputs        = avfilter_vf_vidstabtransform_inputs,
+    .outputs       = avfilter_vf_vidstabtransform_outputs,
+    .priv_class    = &vidstabtransform_class,
+
+};
+

libavfilter/vidstabutils.c

@@ -0,0 +1,84 @@
+/*
+ * Copyright (c) 2013 Georg Martius <georg dot martius at web dot de>
+ *
+ * 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
+ */
+
+#include "vidstabutils.h"
+
+
+/** convert AV's pixelformat to vid.stab pixelformat */
+VSPixelFormat av_2_vs_pixel_format(AVFilterContext *ctx, enum AVPixelFormat pf){
+    switch(pf){
+    case AV_PIX_FMT_YUV420P:  return PF_YUV420P;
+    case AV_PIX_FMT_YUV422P:  return PF_YUV422P;
+    case AV_PIX_FMT_YUV444P:  return PF_YUV444P;
+    case AV_PIX_FMT_YUV410P:  return PF_YUV410P;
+    case AV_PIX_FMT_YUV411P:  return PF_YUV411P;
+    case AV_PIX_FMT_YUV440P:  return PF_YUV440P;
+    case AV_PIX_FMT_YUVA420P: return PF_YUVA420P;
+    case AV_PIX_FMT_GRAY8:    return PF_GRAY8;
+    case AV_PIX_FMT_RGB24:    return PF_RGB24;
+    case AV_PIX_FMT_BGR24:    return PF_BGR24;
+    case AV_PIX_FMT_RGBA:     return PF_RGBA;
+    default:
+        av_log(ctx, AV_LOG_ERROR, "cannot deal with pixel format %i\n", pf);
+        return PF_NONE;
+    }
+}
+
+
+/** struct to hold a valid context for logging from within vid.stab lib */
+typedef struct {
+    const AVClass* class;
+} VS2AVLogCtx;
+
+/** wrapper to log vs_log into av_log */
+static int vs_2_av_log_wrapper(int type, const char* tag, const char* format, ...){
+    va_list ap;
+    VS2AVLogCtx ctx;
+    AVClass class = {
+      .class_name = tag,
+      .item_name  = av_default_item_name,
+      .option     = 0,
+      .version    = LIBAVUTIL_VERSION_INT,
+      .category   = AV_CLASS_CATEGORY_FILTER,
+    };
+    ctx.class = &class;
+    va_start (ap, format);
+    av_vlog(&ctx, type, format, ap);
+    va_end (ap);
+    return VS_OK;
+}
+
+/** sets the memory allocation function and logging constants to av versions */
+void vs_set_mem_and_log_functions(void){
+    vs_malloc  = av_malloc;
+    vs_zalloc  = av_mallocz;
+    vs_realloc = av_realloc;
+    vs_free    = av_free;
+
+    VS_ERROR_TYPE = AV_LOG_ERROR;
+    VS_WARN_TYPE  = AV_LOG_WARNING;
+    VS_INFO_TYPE  = AV_LOG_INFO;
+    VS_MSG_TYPE   = AV_LOG_VERBOSE;
+
+    vs_log   = vs_2_av_log_wrapper;
+
+    VS_ERROR = 0;
+    VS_OK    = 1;
+}

libavfilter/vidstabutils.h

@@ -0,0 +1,36 @@
+/*
+ * Copyright (c) 2013 Georg Martius <georg dot martius at web dot de>
+ *
+ * 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
+ */
+
+#ifndef AVFILTER_VIDSTABUTILS_H
+#define AVFILTER_VIDSTABUTILS_H
+
+#include <vid.stab/libvidstab.h>
+
+#include "avfilter.h"
+
+/* ** some conversions from avlib to vid.stab constants and functions *** */
+
+/** converts the pixelformat of avlib into the one of the vid.stab library */
+VSPixelFormat av_2_vs_pixel_format(AVFilterContext *ctx, enum AVPixelFormat pf);
+
+/** sets the memory allocation function and logging constants to av versions */
+void vs_set_mem_and_log_functions(void);
+
+#endif