avfilter: add gblur filter

Signed-off-by: Paul B Mahol <onemda@gmail.com>

4 files changed, 355 insertions, 1 deletions

diff --git a/libavfilter/Makefile b/libavfilter/Makefile
index 8bc6b9ccb2..fc03fdf09a 100644
--- a/libavfilter/Makefile
+++ b/libavfilter/Makefile
@@ -182,6 +182,7 @@ OBJS-$(CONFIG_FRAMERATE_FILTER)              += vf_framerate.o
 OBJS-$(CONFIG_FRAMESTEP_FILTER)              += vf_framestep.o
 OBJS-$(CONFIG_FREI0R_FILTER)                 += vf_frei0r.o
 OBJS-$(CONFIG_FSPP_FILTER)                   += vf_fspp.o
+OBJS-$(CONFIG_GBLUR_FILTER)                  += vf_gblur.o
 OBJS-$(CONFIG_GEQ_FILTER)                    += vf_geq.o
 OBJS-$(CONFIG_GRADFUN_FILTER)                += vf_gradfun.o
 OBJS-$(CONFIG_HALDCLUT_FILTER)               += vf_lut3d.o dualinput.o framesync.o
diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
index a651bb5072..ac8b41d348 100644
--- a/libavfilter/allfilters.c
+++ b/libavfilter/allfilters.c
@@ -199,6 +199,7 @@ void avfilter_register_all(void)
     REGISTER_FILTER(FRAMESTEP,      framestep,      vf);
     REGISTER_FILTER(FREI0R,         frei0r,         vf);
     REGISTER_FILTER(FSPP,           fspp,           vf);
+    REGISTER_FILTER(GBLUR,          gblur,          vf);
     REGISTER_FILTER(GEQ,            geq,            vf);
     REGISTER_FILTER(GRADFUN,        gradfun,        vf);
     REGISTER_FILTER(HALDCLUT,       haldclut,       vf);
diff --git a/libavfilter/version.h b/libavfilter/version.h
index 5b5f75f74a..6b59d1f222 100644
--- a/libavfilter/version.h
+++ b/libavfilter/version.h
@@ -30,7 +30,7 @@
 #include "libavutil/version.h"
 
 #define LIBAVFILTER_VERSION_MAJOR   6
-#define LIBAVFILTER_VERSION_MINOR  59
+#define LIBAVFILTER_VERSION_MINOR  60
 #define LIBAVFILTER_VERSION_MICRO 100
 
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
diff --git a/libavfilter/vf_gblur.c b/libavfilter/vf_gblur.c
new file mode 100644
index 0000000000..e289f6933f
--- /dev/null
+++ b/libavfilter/vf_gblur.c
@@ -0,0 +1,352 @@
+/*
+ * Copyright (c) 2011 Pascal Getreuer
+ * Copyright (c) 2016 Paul B Mahol
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ *  * Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  * Redistributions in binary form must reproduce the above
+ *    copyright notice, this list of conditions and the following
+ *    disclaimer in the documentation and/or other materials provided
+ *    with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * HOLDER BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "libavutil/imgutils.h"
+#include "libavutil/opt.h"
+#include "libavutil/pixdesc.h"
+#include "avfilter.h"
+#include "formats.h"
+#include "internal.h"
+#include "video.h"
+
+typedef struct GBlurContext {
+    const AVClass *class;
+
+    float sigma;
+    int steps;
+    int planes;
+
+    int depth;
+    int planewidth[4];
+    int planeheight[4];
+    float *buffer;
+    float boundaryscale;
+    float postscale;
+    float nu;
+    int nb_planes;
+} GBlurContext;
+
+#define OFFSET(x) offsetof(GBlurContext, x)
+#define FLAGS AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
+
+static const AVOption gblur_options[] = {
+    { "sigma",  "set sigma",            OFFSET(sigma),  AV_OPT_TYPE_FLOAT, {.dbl=0.5}, 0.0, 1024, FLAGS },
+    { "steps",  "set number of steps",  OFFSET(steps),  AV_OPT_TYPE_INT,   {.i64=1},     1,    6, FLAGS },
+    { "planes", "set planes to filter", OFFSET(planes), AV_OPT_TYPE_INT,   {.i64=0xF},   0,  0xF, FLAGS },
+    { NULL }
+};
+
+AVFILTER_DEFINE_CLASS(gblur);
+
+typedef struct ThreadData {
+    int height;
+    int width;
+} ThreadData;
+
+static int filter_horizontally(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
+{
+    GBlurContext *s = ctx->priv;
+    ThreadData *td = arg;
+    const int height = td->height;
+    const int width = td->width;
+    const int slice_start = (height *  jobnr   ) / nb_jobs;
+    const int slice_end   = (height * (jobnr+1)) / nb_jobs;
+    const float boundaryscale = s->boundaryscale;
+    const int steps = s->steps;
+    const float nu = s->nu;
+    float *buffer = s->buffer;
+    int y, x, step;
+    float *ptr;
+
+    /* Filter horizontally along each row */
+    for (y = slice_start; y < slice_end; y++) {
+        for (step = 0; step < steps; step++) {
+            ptr = buffer + width * y;
+            ptr[0] *= boundaryscale;
+
+            /* Filter rightwards */
+            for (x = 1; x < width; x++)
+                ptr[x] += nu * ptr[x - 1];
+
+            ptr[x = width - 1] *= boundaryscale;
+
+            /* Filter leftwards */
+            for (; x > 0; x--)
+                ptr[x - 1] += nu * ptr[x];
+        }
+    }
+
+    return 0;
+}
+
+static int filter_vertically(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
+{
+    GBlurContext *s = ctx->priv;
+    ThreadData *td = arg;
+    const int height = td->height;
+    const int width = td->width;
+    const int slice_start = (width *  jobnr   ) / nb_jobs;
+    const int slice_end   = (width * (jobnr+1)) / nb_jobs;
+    const float boundaryscale = s->boundaryscale;
+    const int numpixels = width * height;
+    const int steps = s->steps;
+    const float nu = s->nu;
+    float *buffer = s->buffer;
+    int i, x, step;
+    float *ptr;
+
+    /* Filter vertically along each column */
+    for (x = slice_start; x < slice_end; x++) {
+        for (step = 0; step < steps; step++) {
+            ptr = buffer + x;
+            ptr[0] *= boundaryscale;
+
+            /* Filter downwards */
+            for (i = width; i < numpixels; i += width)
+                ptr[i] += nu * ptr[i - width];
+
+            ptr[i = numpixels - width] *= boundaryscale;
+
+            /* Filter upwards */
+            for (; i > 0; i -= width)
+                ptr[i - width] += nu * ptr[i];
+        }
+    }
+
+    return 0;
+}
+
+
+static int filter_postscale(AVFilterContext *ctx, void *arg, int jobnr, int nb_jobs)
+{
+    GBlurContext *s = ctx->priv;
+    ThreadData *td = arg;
+    const int height = td->height;
+    const int width = td->width;
+    const int64_t numpixels = width * height;
+    const int slice_start = (numpixels *  jobnr   ) / nb_jobs;
+    const int slice_end   = (numpixels * (jobnr+1)) / nb_jobs;
+    const float postscale = s->postscale;
+    float *buffer = s->buffer;
+    int i;
+
+    for (i = slice_start; i < slice_end; i++)
+        buffer[i] *= postscale;
+
+    return 0;
+}
+
+static void gaussianiir2d(AVFilterContext *ctx, int plane)
+{
+    GBlurContext *s = ctx->priv;
+    const int width = s->planewidth[plane];
+    const int height = s->planeheight[plane];
+    const int nb_threads = ff_filter_get_nb_threads(ctx);
+    ThreadData td;
+
+    if (s->sigma <= 0 || s->steps < 0)
+        return;
+
+    td.width = width;
+    td.height = height;
+    ctx->internal->execute(ctx, filter_horizontally, &td, NULL, FFMIN(height, nb_threads));
+    ctx->internal->execute(ctx, filter_vertically, &td, NULL, FFMIN(width, nb_threads));
+    ctx->internal->execute(ctx, filter_postscale, &td, NULL, FFMIN(width * height, nb_threads));
+}
+
+static int query_formats(AVFilterContext *ctx)
+{
+    static const enum AVPixelFormat pix_fmts[] = {
+        AV_PIX_FMT_YUVA444P, AV_PIX_FMT_YUV444P, AV_PIX_FMT_YUV440P,
+        AV_PIX_FMT_YUVJ444P, AV_PIX_FMT_YUVJ440P,
+        AV_PIX_FMT_YUVA422P, AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUVA420P, AV_PIX_FMT_YUV420P,
+        AV_PIX_FMT_YUVJ422P, AV_PIX_FMT_YUVJ420P,
+        AV_PIX_FMT_YUVJ411P, AV_PIX_FMT_YUV411P, AV_PIX_FMT_YUV410P,
+        AV_PIX_FMT_YUV420P9, AV_PIX_FMT_YUV422P9, AV_PIX_FMT_YUV444P9,
+        AV_PIX_FMT_YUV420P10, AV_PIX_FMT_YUV422P10, AV_PIX_FMT_YUV444P10,
+        AV_PIX_FMT_YUV420P12, AV_PIX_FMT_YUV422P12, AV_PIX_FMT_YUV444P12, AV_PIX_FMT_YUV440P12,
+        AV_PIX_FMT_YUV420P14, AV_PIX_FMT_YUV422P14, AV_PIX_FMT_YUV444P14,
+        AV_PIX_FMT_YUV420P16, AV_PIX_FMT_YUV422P16, AV_PIX_FMT_YUV444P16,
+        AV_PIX_FMT_YUVA420P9, AV_PIX_FMT_YUVA422P9, AV_PIX_FMT_YUVA444P9,
+        AV_PIX_FMT_YUVA420P10, AV_PIX_FMT_YUVA422P10, AV_PIX_FMT_YUVA444P10,
+        AV_PIX_FMT_YUVA420P16, AV_PIX_FMT_YUVA422P16, AV_PIX_FMT_YUVA444P16,
+        AV_PIX_FMT_GBRP, AV_PIX_FMT_GBRP9, AV_PIX_FMT_GBRP10,
+        AV_PIX_FMT_GBRP12, AV_PIX_FMT_GBRP14, AV_PIX_FMT_GBRP16,
+        AV_PIX_FMT_GBRAP, AV_PIX_FMT_GBRAP12, AV_PIX_FMT_GBRAP16,
+        AV_PIX_FMT_GRAY8, AV_PIX_FMT_GRAY16,
+        AV_PIX_FMT_NONE
+    };
+
+    return ff_set_common_formats(ctx, ff_make_format_list(pix_fmts));
+}
+
+static int config_input(AVFilterLink *inlink)
+{
+    const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
+    GBlurContext *s = inlink->dst->priv;
+
+    s->depth = desc->comp[0].depth;
+    s->planewidth[1] = s->planewidth[2] = AV_CEIL_RSHIFT(inlink->w, desc->log2_chroma_w);
+    s->planewidth[0] = s->planewidth[3] = inlink->w;
+    s->planeheight[1] = s->planeheight[2] = AV_CEIL_RSHIFT(inlink->h, desc->log2_chroma_h);
+    s->planeheight[0] = s->planeheight[3] = inlink->h;
+
+    s->nb_planes = av_pix_fmt_count_planes(inlink->format);
+
+    s->buffer = av_malloc_array(inlink->w, inlink->h * sizeof(*s->buffer));
+    if (!s->buffer)
+        return AVERROR(ENOMEM);
+    return 0;
+}
+
+static int filter_frame(AVFilterLink *inlink, AVFrame *in)
+{
+    AVFilterContext *ctx = inlink->dst;
+    GBlurContext *s = ctx->priv;
+    AVFilterLink *outlink = ctx->outputs[0];
+    double dnu, lambda;
+    float sigma = s->sigma;
+    int steps = s->steps;
+    AVFrame *out;
+    int plane;
+
+    lambda = (sigma * sigma) / (2.0 * steps);
+    dnu = (1.0 + 2.0 * lambda - sqrt(1.0 + 4.0 * lambda)) / (2.0 * lambda);
+    s->postscale = pow(dnu / lambda, 2 * steps);
+    s->boundaryscale = 1.0 / (1.0 - dnu);
+    s->nu = (float)dnu;
+
+    if (av_frame_is_writable(in)) {
+        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 < s->nb_planes; plane++) {
+        const int height = s->planeheight[plane];
+        const int width = s->planewidth[plane];
+        float *bptr = s->buffer;
+        const uint8_t *src = in->data[plane];
+        const uint16_t *src16 = (const uint16_t *)in->data[plane];
+        uint8_t *dst = out->data[plane];
+        uint16_t *dst16 = (uint16_t *)out->data[plane];
+        int y, x;
+
+        if (!s->sigma || !(s->planes & (1 << plane))) {
+            if (out != in)
+                av_image_copy_plane(out->data[plane], out->linesize[plane],
+                                    in->data[plane], in->linesize[plane],
+                                    width * ((s->depth + 7) / 8), height);
+            continue;
+        }
+
+        if (s->depth == 8) {
+            for (y = 0; y < height; y++) {
+                for (x = 0; x < width; x++) {
+                    bptr[x] = src[x];
+                }
+                bptr += width;
+                src += in->linesize[plane];
+            }
+        } else {
+            for (y = 0; y < height; y++) {
+                for (x = 0; x < width; x++) {
+                    bptr[x] = src16[x];
+                }
+                bptr += width;
+                src16 += in->linesize[plane] / 2;
+            }
+        }
+
+        gaussianiir2d(ctx, plane);
+
+        bptr = s->buffer;
+        if (s->depth == 8) {
+            for (y = 0; y < height; y++) {
+                for (x = 0; x < width; x++) {
+                    dst[x] = bptr[x];
+                }
+                bptr += width;
+                dst += out->linesize[plane];
+            }
+        } else {
+            for (y = 0; y < height; y++) {
+                for (x = 0; x < width; x++) {
+                    dst16[x] = bptr[x];
+                }
+                bptr += width;
+                dst16 += out->linesize[plane] / 2;
+            }
+        }
+    }
+
+    if (out != in)
+        av_frame_free(&in);
+    return ff_filter_frame(outlink, out);
+}
+
+static av_cold void uninit(AVFilterContext *ctx)
+{
+    GBlurContext *s = ctx->priv;
+
+    av_freep(&s->buffer);
+}
+
+static const AVFilterPad gblur_inputs[] = {
+    {
+        .name         = "default",
+        .type         = AVMEDIA_TYPE_VIDEO,
+        .config_props = config_input,
+        .filter_frame = filter_frame,
+    },
+    { NULL }
+};
+
+static const AVFilterPad gblur_outputs[] = {
+    {
+        .name = "default",
+        .type = AVMEDIA_TYPE_VIDEO,
+    },
+    { NULL }
+};
+
+AVFilter ff_vf_gblur = {
+    .name          = "gblur",
+    .description   = NULL_IF_CONFIG_SMALL("Apply Gaussian Blur filter."),
+    .priv_size     = sizeof(GBlurContext),
+    .priv_class    = &gblur_class,
+    .uninit        = uninit,
+    .query_formats = query_formats,
+    .inputs        = gblur_inputs,
+    .outputs       = gblur_outputs,
+    .flags         = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC | AVFILTER_FLAG_SLICE_THREADS,
+};