avfilter: add readvitc filter

Add a filter to scan the top lines of video frames for vertical interval
timecode (VITC) information and attach it as metadata keys.

Signed-off-by: Tobias Rapp <t.rapp@noa-archive.com>

7 files changed, 300 insertions, 1 deletions

diff --git a/Changelog b/Changelog
index 18fdb27d39..6e40739cc9 100644
--- a/Changelog
+++ b/Changelog
@@ -25,6 +25,7 @@ version <next>:
 - hash and framehash muxers
 - colorspace filter
 - hdcd filter
+- readvitc filter
 
 version 3.0:
 - Common Encryption (CENC) MP4 encoding and decoding support
diff --git a/MAINTAINERS b/MAINTAINERS
index a993a675a2..d2593e52c1 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -385,6 +385,7 @@ Filters:
   vf_neighbor.c                         Paul B Mahol
   vf_psnr.c                             Paul B Mahol
   vf_random.c                           Paul B Mahol
+  vf_readvitc.c                         Tobias Rapp (CC t.rapp at noa-archive dot com)
   vf_scale.c                            Michael Niedermayer
   vf_separatefields.c                   Paul B Mahol
   vf_ssim.c                             Paul B Mahol
diff --git a/doc/filters.texi b/doc/filters.texi
index 61307e5838..b17b115288 100644
--- a/doc/filters.texi
+++ b/doc/filters.texi
@@ -10618,6 +10618,43 @@ less than @code{0}, the filter will try to use a good random seed on a
 best effort basis.
 @end table
 
+@section readvitc
+
+Read vertical interval timecode (VITC) information from the top lines of a
+video frame.
+
+The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
+timecode value, if a valid timecode has been detected. Further metadata key
+@code{lavfi.readvitc.found} is set to 0/1 depending on whether
+timecode data has been found or not.
+
+This filter accepts the following options:
+
+@table @option
+@item scan_max
+Set the maximum number of lines to scan for VITC data. If the value is set to
+@code{-1} the full video frame is scanned. Default is @code{45}.
+
+@item thr_b
+Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
+default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
+
+@item thr_w
+Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
+default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
+@end table
+
+@subsection Examples
+
+@itemize
+@item
+Detect and draw VITC data onto the video frame; if no valid VITC is detected,
+draw @code{--:--:--:--} as a placeholder:
+@example
+ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
+@end example
+@end itemize
+
 @section remap
 
 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
diff --git a/libavfilter/Makefile b/libavfilter/Makefile
index 2ea8c84a0d..f1f37cb093 100644
--- a/libavfilter/Makefile
+++ b/libavfilter/Makefile
@@ -227,6 +227,7 @@ OBJS-$(CONFIG_PSNR_FILTER)                   += vf_psnr.o dualinput.o framesync.
 OBJS-$(CONFIG_PULLUP_FILTER)                 += vf_pullup.o
 OBJS-$(CONFIG_QP_FILTER)                     += vf_qp.o
 OBJS-$(CONFIG_RANDOM_FILTER)                 += vf_random.o
+OBJS-$(CONFIG_READVITC_FILTER)               += vf_readvitc.o
 OBJS-$(CONFIG_REALTIME_FILTER)               += f_realtime.o
 OBJS-$(CONFIG_REMAP_FILTER)                  += vf_remap.o framesync.o
 OBJS-$(CONFIG_REMOVEGRAIN_FILTER)            += vf_removegrain.o
diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
index 52d3f619ca..ad6ba21da5 100644
--- a/libavfilter/allfilters.c
+++ b/libavfilter/allfilters.c
@@ -247,6 +247,7 @@ void avfilter_register_all(void)
     REGISTER_FILTER(PULLUP,         pullup,         vf);
     REGISTER_FILTER(QP,             qp,             vf);
     REGISTER_FILTER(RANDOM,         random,         vf);
+    REGISTER_FILTER(READVITC,       readvitc,       vf);
     REGISTER_FILTER(REALTIME,       realtime,       vf);
     REGISTER_FILTER(REMAP,          remap,          vf);
     REGISTER_FILTER(REMOVEGRAIN,    removegrain,    vf);
diff --git a/libavfilter/version.h b/libavfilter/version.h
index 48b9d9a6e5..f4ade1cf15 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  42
+#define LIBAVFILTER_VERSION_MINOR  43
 #define LIBAVFILTER_VERSION_MICRO 100
 
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
diff --git a/libavfilter/vf_readvitc.c b/libavfilter/vf_readvitc.c
new file mode 100644
index 0000000000..7ef8cdae58
--- /dev/null
+++ b/libavfilter/vf_readvitc.c
@@ -0,0 +1,258 @@
+/*
+ * Copyright (c) 2016 Tobias Rapp
+ *
+ * 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
+ * Filter for reading the vertical interval timecode (VITC).
+ * See also https://en.wikipedia.org/wiki/Vertical_interval_timecode
+ */
+
+#include "libavutil/common.h"
+#include "libavutil/internal.h"
+#include "libavutil/opt.h"
+#include "libavutil/pixdesc.h"
+#include "libavutil/timecode.h"
+#include "avfilter.h"
+#include "formats.h"
+#include "internal.h"
+
+#define LINE_DATA_SIZE 9
+
+typedef struct ReadVitcContext {
+    const AVClass *class;
+
+    int scan_max;
+    double thr_b;
+    double thr_w;
+
+    int threshold_black;
+    int threshold_white;
+    int threshold_gray;
+    int grp_width;
+    uint8_t line_data[LINE_DATA_SIZE];
+    char tcbuf[AV_TIMECODE_STR_SIZE];
+} ReadVitcContext;
+
+#define OFFSET(x) offsetof(ReadVitcContext, x)
+#define FLAGS AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
+
+static const AVOption readvitc_options[] = {
+    { "scan_max", "maximum line numbers to scan for VITC data", OFFSET(scan_max), AV_OPT_TYPE_INT, {.i64 = 45 }, -1, INT_MAX, FLAGS },
+    { "thr_b",    "black color threshold", OFFSET(thr_b), AV_OPT_TYPE_DOUBLE, {.dbl = 0.2 }, 0, 1.0, FLAGS },
+    { "thr_w",    "white color threshold", OFFSET(thr_w), AV_OPT_TYPE_DOUBLE, {.dbl = 0.6 }, 0, 1.0, FLAGS },
+    { NULL }
+};
+
+AVFILTER_DEFINE_CLASS(readvitc);
+
+static uint8_t get_vitc_crc( uint8_t *line ) {
+    uint8_t crc;
+
+    crc = 0x01 | (line[0] << 2);
+    crc ^= (line[0] >> 6) | 0x04 | (line[1] << 4);
+    crc ^= (line[1] >> 4) | 0x10 | (line[2] << 6);
+    crc ^= (line[2] >> 2) | 0x40;
+    crc ^= line[3];
+    crc ^= 0x01 | (line[4] << 2);
+    crc ^= (line[4] >> 6) | 0x04 | (line[5] << 4);
+    crc ^= (line[5] >> 4) | 0x10 | (line[6] << 6);
+    crc ^= (line[6] >> 2) | 0x40;
+    crc ^= line[7];
+    crc ^= 0x01;
+    crc = (crc >> 2) | (crc << 6);  // rotate byte right by two bits
+    return crc;
+}
+
+static inline uint8_t get_pit_avg3( uint8_t *line, int i ) {
+    return ((line[i-1] + line[i] + line[i+1]) / 3);
+}
+
+static int read_vitc_line( ReadVitcContext *ctx, uint8_t *src, int line_size, int width, int height )
+{
+    uint8_t *scan_line;
+    int grp_index, pit_index;
+    int grp_start_pos;
+    uint8_t pit_value;
+    int x, y, res = 0;
+
+    if (ctx->scan_max >= 0)
+        height = FFMIN(height, ctx->scan_max);
+
+    // scan lines for VITC data, starting from the top
+    for (y = 0; y < height; y++) {
+        scan_line = src;
+        memset(ctx->line_data, 0, LINE_DATA_SIZE);
+        grp_index = 0;
+        x = 0;
+        while ((x < width) && (grp_index < 9)) {
+            // search next sync pattern
+            while ((x < width) && (scan_line[x] < ctx->threshold_white))
+                x++;
+            while ((x < width) && (scan_line[x] > ctx->threshold_black))
+                x++;
+            x = FFMAX(x - ((ctx->grp_width+10) / 20), 1);  // step back a half pit
+            grp_start_pos = x;
+            if ((grp_start_pos + ctx->grp_width) > width)
+                break;  // not enough pixels for reading a whole pit group
+            pit_value = get_pit_avg3(scan_line, x);
+            if (pit_value < ctx->threshold_white)
+               break;  // first sync bit mismatch
+            x = grp_start_pos + ((ctx->grp_width) / 10);
+            pit_value = get_pit_avg3(scan_line, x);
+            if (pit_value > ctx->threshold_black )
+                break;  // second sync bit mismatch
+            for (pit_index = 0; pit_index <= 7; pit_index++) {
+                x = grp_start_pos + (((pit_index+2)*ctx->grp_width) / 10);
+                pit_value = get_pit_avg3(scan_line, x);
+                if (pit_value > ctx->threshold_gray)
+                    ctx->line_data[grp_index] |= (1 << pit_index);
+            }
+            grp_index++;
+        }
+        if ((grp_index == 9) && (get_vitc_crc(ctx->line_data) == ctx->line_data[8])) {
+            res = 1;
+            break;
+        }
+        src += line_size;
+    }
+
+    return res;
+}
+
+static unsigned bcd2uint(uint8_t high, uint8_t low)
+{
+   if (high > 9 || low > 9)
+       return 0;
+   return 10*high + low;
+}
+
+static char *make_vitc_tc_string(char *buf, uint8_t *line)
+{
+    unsigned hh   = bcd2uint(line[7] & 0x03, line[6] & 0x0f);  // 6-bit hours
+    unsigned mm   = bcd2uint(line[5] & 0x07, line[4] & 0x0f);  // 7-bit minutes
+    unsigned ss   = bcd2uint(line[3] & 0x07, line[2] & 0x0f);  // 7-bit seconds
+    unsigned ff   = bcd2uint(line[1] & 0x03, line[0] & 0x0f);  // 6-bit frames
+    unsigned drop = (line[1] & 0x04);                          // 1-bit drop flag
+    snprintf(buf, AV_TIMECODE_STR_SIZE, "%02u:%02u:%02u%c%02u",
+             hh, mm, ss, drop ? ';' : ':', ff);
+    return buf;
+}
+
+static av_cold int init(AVFilterContext *ctx)
+{
+    ReadVitcContext *s = ctx->priv;
+
+    s->threshold_black = s->thr_b * UINT8_MAX;
+    s->threshold_white = s->thr_w * UINT8_MAX;
+    if (s->threshold_black > s->threshold_white) {
+        av_log(ctx, AV_LOG_WARNING, "Black color threshold is higher than white color threshold (%g > %g)\n",
+                s->thr_b, s->thr_w);
+        return AVERROR(EINVAL);
+    }
+    s->threshold_gray = s->threshold_white - ((s->threshold_white - s->threshold_black) / 2);
+    av_log(ctx, AV_LOG_DEBUG, "threshold_black:%d threshold_white:%d threshold_gray:%d\n",
+            s->threshold_black, s->threshold_white, s->threshold_gray);
+
+    return 0;
+}
+
+static int config_props(AVFilterLink *inlink)
+{
+    AVFilterContext *ctx = inlink->dst;
+    ReadVitcContext *s = ctx->priv;
+
+    s->grp_width = inlink->w * 5 / 48;
+    av_log(ctx, AV_LOG_DEBUG, "w:%d h:%d grp_width:%d scan_max:%d\n",
+            inlink->w, inlink->h, s->grp_width, s->scan_max);
+    return 0;
+}
+
+static int query_formats(AVFilterContext *ctx)
+{
+    static const enum AVPixelFormat pixel_fmts[] = {
+        AV_PIX_FMT_GRAY8,
+        AV_PIX_FMT_NV12,
+        AV_PIX_FMT_NV16,
+        AV_PIX_FMT_NV21,
+        AV_PIX_FMT_YUV410P,
+        AV_PIX_FMT_YUV411P,
+        AV_PIX_FMT_YUV420P,
+        AV_PIX_FMT_YUV422P,
+        AV_PIX_FMT_YUV440P,
+        AV_PIX_FMT_YUV444P,
+        AV_PIX_FMT_YUVA420P,
+        AV_PIX_FMT_YUVA422P,
+        AV_PIX_FMT_YUVA444P,
+        AV_PIX_FMT_YUVJ411P,
+        AV_PIX_FMT_YUVJ420P,
+        AV_PIX_FMT_YUVJ422P,
+        AV_PIX_FMT_YUVJ440P,
+        AV_PIX_FMT_YUVJ444P,
+        AV_PIX_FMT_NONE
+    };
+    AVFilterFormats *fmts_list = ff_make_format_list(pixel_fmts);
+    if (!fmts_list)
+        return AVERROR(ENOMEM);
+    return ff_set_common_formats(ctx, fmts_list);
+}
+
+static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
+{
+    AVFilterContext *ctx = inlink->dst;
+    AVFilterLink *outlink = ctx->outputs[0];
+    ReadVitcContext *s = ctx->priv;
+    int found;
+
+    found = read_vitc_line(s, frame->data[0], frame->linesize[0], inlink->w, inlink->h);
+    av_dict_set(&frame->metadata, "lavfi.readvitc.found", (found ? "1" : "0"), 0);
+    if (found)
+        av_dict_set(&frame->metadata, "lavfi.readvitc.tc_str", make_vitc_tc_string(s->tcbuf, s->line_data), 0);
+
+    return ff_filter_frame(outlink, frame);
+}
+
+static const AVFilterPad inputs[] = {
+    {
+        .name         = "default",
+        .type         = AVMEDIA_TYPE_VIDEO,
+        .filter_frame = filter_frame,
+        .config_props = config_props,
+    },
+    { NULL }
+};
+
+static const AVFilterPad outputs[] = {
+    {
+        .name = "default",
+        .type = AVMEDIA_TYPE_VIDEO,
+    },
+    { NULL }
+};
+
+AVFilter ff_vf_readvitc = {
+    .name          = "readvitc",
+    .description   = NULL_IF_CONFIG_SMALL("Read vertical interval timecode and write it to frame metadata."),
+    .priv_size     = sizeof(ReadVitcContext),
+    .priv_class    = &readvitc_class,
+    .inputs        = inputs,
+    .outputs       = outputs,
+    .init          = init,
+    .query_formats = query_formats,
+};