gstreamer/subprojects/gst-plugins-bad/ext/opencv/gstskindetect.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

395 lines
14 KiB
C++
Raw Normal View History

/*
* GStreamer
* Copyright (C) 2013 Miguel Casas-Sanchez <miguelecasassanchez@gmail.com>
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
*
* Alternatively, the contents of this file may be used under the
* GNU Lesser General Public License Version 2.1 (the "LGPL"), in
* which case the following provisions apply instead of the ones
* mentioned above:
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library 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
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public
* License along with this library; if not, write to the
* Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
* Boston, MA 02110-1301, USA.
*/
/**
* SECTION:element-skindetect
*
* Human skin detection on videos and images
*
2019-05-29 20:58:08 +00:00
* ## Example launch line
*
* |[
* gst-launch-1.0 videotestsrc ! decodebin ! videoconvert ! skindetect ! videoconvert ! xvimagesink
* ]|
*/
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif
#include "gstskindetect.h"
2018-08-02 15:03:47 +00:00
#include <opencv2/imgproc.hpp>
GST_DEBUG_CATEGORY_STATIC (gst_skin_detect_debug);
#define GST_CAT_DEFAULT gst_skin_detect_debug
/* Filter signals and args */
enum
{
/* FILL ME */
LAST_SIGNAL
};
enum
{
PROP_0,
PROP_POSTPROCESS,
PROP_METHOD,
PROP_MASK
};
typedef enum
{
HSV,
RGB
} GstSkindetectMethod;
#define GST_TYPE_SKIN_DETECT_METHOD (gst_skin_detect_method_get_type ())
static GType
gst_skin_detect_method_get_type (void)
{
static GType etype = 0;
if (etype == 0) {
static const GEnumValue values[] = {
{HSV, "Classic HSV thresholding", "hsv"},
{RGB, "Normalised-RGB colorspace thresholding", "rgb"},
{0, NULL, NULL},
};
etype = g_enum_register_static ("GstSkindetectMethod", values);
}
return etype;
}
G_DEFINE_TYPE_WITH_CODE (GstSkinDetect, gst_skin_detect,
GST_TYPE_OPENCV_VIDEO_FILTER,
GST_DEBUG_CATEGORY_INIT (gst_skin_detect_debug, "skindetect", 0,
"Performs skin detection on videos and images");
);
GST_ELEMENT_REGISTER_DEFINE (skindetect, "skindetect", GST_RANK_NONE,
GST_TYPE_SKIN_DETECT);
static GstStaticPadTemplate sink_factory = GST_STATIC_PAD_TEMPLATE ("sink",
GST_PAD_SINK,
GST_PAD_ALWAYS,
GST_STATIC_CAPS (GST_VIDEO_CAPS_MAKE ("RGB")));
static GstStaticPadTemplate src_factory = GST_STATIC_PAD_TEMPLATE ("src",
GST_PAD_SRC,
GST_PAD_ALWAYS,
GST_STATIC_CAPS (GST_VIDEO_CAPS_MAKE ("RGB")));
static void gst_skin_detect_set_property (GObject * object, guint prop_id,
const GValue * value, GParamSpec * pspec);
static void gst_skin_detect_get_property (GObject * object, guint prop_id,
GValue * value, GParamSpec * pspec);
static GstFlowReturn gst_skin_detect_transform (GstOpencvVideoFilter * filter,
2018-12-01 21:48:53 +00:00
GstBuffer * buf, cv::Mat img, GstBuffer * outbuf, cv::Mat outimg);
2018-12-01 21:48:53 +00:00
static void gst_skin_detect_finalize (GObject * object);
static gboolean
gst_skin_detect_set_caps (GstOpencvVideoFilter * transform,
2018-12-01 21:48:53 +00:00
gint in_width, gint in_height, int in_cv_type,
gint out_width, gint out_height, int out_cv_type);
/* initialize the skindetect's class */
static void
gst_skin_detect_class_init (GstSkinDetectClass * klass)
{
GObjectClass *gobject_class;
GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
GstOpencvVideoFilterClass *gstopencvbasefilter_class;
gobject_class = (GObjectClass *) klass;
gstopencvbasefilter_class = (GstOpencvVideoFilterClass *) klass;
2018-12-01 21:48:53 +00:00
gobject_class->finalize = GST_DEBUG_FUNCPTR (gst_skin_detect_finalize);
gobject_class->set_property = gst_skin_detect_set_property;
gobject_class->get_property = gst_skin_detect_get_property;
gstopencvbasefilter_class->cv_trans_func = gst_skin_detect_transform;
g_object_class_install_property (gobject_class, PROP_POSTPROCESS,
g_param_spec_boolean ("postprocess", "Postprocess",
"Apply opening-closing to skin detection to extract large, significant blobs ",
TRUE, (GParamFlags)
(G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS)));
g_object_class_install_property (gobject_class, PROP_METHOD,
g_param_spec_enum ("method",
"Method to use",
"Method to use",
GST_TYPE_SKIN_DETECT_METHOD, HSV,
(GParamFlags) (G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS)));
gst_element_class_set_static_metadata (element_class,
"skindetect",
"Filter/Effect/Video",
"Performs non-parametric skin detection on input",
"Miguel Casas-Sanchez <miguelecasassanchez@gmail.com>");
gst_element_class_add_static_pad_template (element_class, &src_factory);
gst_element_class_add_static_pad_template (element_class, &sink_factory);
gstopencvbasefilter_class->cv_set_caps = gst_skin_detect_set_caps;
gst_type_mark_as_plugin_api (GST_TYPE_SKIN_DETECT_METHOD, (GstPluginAPIFlags) 0);
}
/* initialize the new element
* instantiate pads and add them to element
2019-09-02 19:08:44 +00:00
* set pad callback functions
* initialize instance structure
*/
static void
gst_skin_detect_init (GstSkinDetect * filter)
{
filter->postprocess = TRUE;
filter->method = HSV;
gst_opencv_video_filter_set_in_place (GST_OPENCV_VIDEO_FILTER_CAST (filter),
FALSE);
}
static void
gst_skin_detect_set_property (GObject * object, guint prop_id,
const GValue * value, GParamSpec * pspec)
{
GstSkinDetect *filter = GST_SKIN_DETECT (object);
switch (prop_id) {
case PROP_POSTPROCESS:
filter->postprocess = g_value_get_boolean (value);
break;
case PROP_METHOD:
filter->method = g_value_get_enum (value);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
gst_skin_detect_get_property (GObject * object, guint prop_id,
GValue * value, GParamSpec * pspec)
{
GstSkinDetect *filter = GST_SKIN_DETECT (object);
switch (prop_id) {
case PROP_POSTPROCESS:
g_value_set_boolean (value, filter->postprocess);
break;
case PROP_METHOD:
g_value_set_enum (value, filter->method);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
/* GstElement vmethod implementations */
/* this function handles the link with other elements */
static gboolean
gst_skin_detect_set_caps (GstOpencvVideoFilter * transform,
2018-12-01 21:48:53 +00:00
gint in_width, gint in_height, int in_cv_type,
gint out_width, gint out_height, int out_cv_type)
{
GstSkinDetect *filter = GST_SKIN_DETECT (transform);
2018-12-01 21:48:53 +00:00
cv::Size size = cv::Size (in_width, in_height);
2018-12-01 21:48:53 +00:00
filter->cvRGB.create (size, CV_8UC3);
filter->cvChA.create (size, CV_8UC1);
filter->width = in_width;
filter->height = in_height;
2018-12-01 21:48:53 +00:00
filter->cvHSV.create (size, CV_8UC3);
filter->cvH.create (size, CV_8UC1); /* Hue component. */
filter->cvH2.create (size, CV_8UC1); /* Hue component, 2nd threshold */
filter->cvS.create (size, CV_8UC1); /* Saturation component. */
filter->cvV.create (size, CV_8UC1); /* Brightness component. */
filter->cvSkinPixels1.create (size, CV_8UC1); /* Greyscale output image */
filter->cvR.create (size, CV_8UC1); /* R component. */
filter->cvG.create (size, CV_8UC1); /* G component. */
filter->cvB.create (size, CV_8UC1); /* B component. */
filter->cvAll.create (size, CV_32FC1); /* (R+G+B) component. */
filter->cvR2.create (size, CV_32FC1); /* R component, 32bits */
filter->cvRp.create (size, CV_32FC1); /* R' and >0.4 */
filter->cvGp.create (size, CV_32FC1); /* G' and > 0.28 */
filter->cvRp2.create (size, CV_32FC1); /* R' <0.6 */
filter->cvGp2.create (size, CV_32FC1); /* G' <0.4 */
filter->cvSkinPixels2.create (size, CV_32FC1); /* Greyscale output image. */
filter->cvdraft.create (size, CV_8UC1); /* Greyscale output image. */
return TRUE;
}
/* Clean up */
static void
2018-12-01 21:48:53 +00:00
gst_skin_detect_finalize (GObject * object)
{
2018-12-01 21:48:53 +00:00
GstSkinDetect *filter = GST_SKIN_DETECT (object);
filter->cvRGB.release ();
filter->cvChA.release ();
filter->cvHSV.release ();
filter->cvH.release ();
filter->cvH2.release ();
filter->cvS.release ();
filter->cvV.release ();
filter->cvSkinPixels1.release ();
filter->cvR.release ();
filter->cvG.release ();
filter->cvB.release ();
filter->cvAll.release ();
filter->cvR2.release ();
filter->cvRp.release ();
filter->cvGp.release ();
filter->cvRp2.release ();
filter->cvGp2.release ();
filter->cvdraft.release ();
filter->cvSkinPixels2.release ();
G_OBJECT_CLASS (gst_skin_detect_parent_class)->finalize (object);
}
static GstFlowReturn
gst_skin_detect_transform (GstOpencvVideoFilter * base, GstBuffer * buf,
2018-12-01 21:48:53 +00:00
cv::Mat img, GstBuffer * outbuf, cv::Mat outimg)
{
GstSkinDetect *filter = GST_SKIN_DETECT (base);
2018-12-01 21:48:53 +00:00
std::vector < cv::Mat > channels (3);
filter->cvRGB = cv::Mat (img);
/* SKIN COLOUR BLOB DETECTION */
if (HSV == filter->method) {
2018-12-01 21:48:53 +00:00
cv::cvtColor (filter->cvRGB, filter->cvHSV, cv::COLOR_RGB2HSV);
cv::split (filter->cvHSV, channels);
filter->cvH = channels.at (0);
filter->cvS = channels.at (1);
filter->cvV = channels.at (2);
/* Detect which pixels in each of the H, S and V channels are probably skin pixels.
Assume that skin has a Hue between 0 to 18 (out of 180), and Saturation above 50, and Brightness above 80. */
2018-12-01 21:48:53 +00:00
cv::threshold (filter->cvH, filter->cvH2, 10, UCHAR_MAX, cv::THRESH_BINARY); /* (hue > 10) */
cv::threshold (filter->cvH, filter->cvH, 20, UCHAR_MAX, cv::THRESH_BINARY_INV); /* (hue < 20) */
cv::threshold (filter->cvS, filter->cvS, 48, UCHAR_MAX, cv::THRESH_BINARY); /* (sat > 48) */
cv::threshold (filter->cvV, filter->cvV, 80, UCHAR_MAX, cv::THRESH_BINARY); /* (val > 80) */
/* erode the HUE to get rid of noise. */
2018-12-01 21:48:53 +00:00
cv::erode (filter->cvH, filter->cvH, cv::Mat (), cv::Point (-1, -1), 1);
/* Combine all 3 thresholded color components, so that an output pixel will only
be white (255) if the H, S and V pixels were also white.
imageSkin = (hue > 10) ^ (hue < 20) ^ (sat > 48) ^ (val > 80), where ^ mean pixels-wise AND */
2018-12-01 21:48:53 +00:00
cv::bitwise_and (filter->cvH, filter->cvS, filter->cvSkinPixels1);
cv::bitwise_and (filter->cvSkinPixels1, filter->cvH2,
filter->cvSkinPixels1);
cv::bitwise_and (filter->cvSkinPixels1, filter->cvV, filter->cvSkinPixels1);
2018-12-01 21:48:53 +00:00
cv::cvtColor (filter->cvSkinPixels1, filter->cvRGB, cv::COLOR_GRAY2RGB);
} else if (RGB == filter->method) {
2018-12-01 21:48:53 +00:00
cv::split (filter->cvRGB, channels);
filter->cvR = channels.at (0);
filter->cvG = channels.at (1);
filter->cvB = channels.at (2);
cv::add (filter->cvR, filter->cvG, filter->cvAll);
cv::add (filter->cvB, filter->cvAll, filter->cvAll); /* All = R + G + B */
cv::divide (filter->cvR, filter->cvAll, filter->cvRp, 1.0, filter->cvRp.type ()); /* R' = R / ( R + G + B) */
cv::divide (filter->cvG, filter->cvAll, filter->cvGp, 1.0, filter->cvGp.type ()); /* G' = G / ( R + G + B) */
filter->cvR.convertTo (filter->cvR2, filter->cvR2.type (), 1.0, 0.0);
filter->cvGp.copyTo (filter->cvGp2);
filter->cvRp.copyTo (filter->cvRp2);
cv::threshold (filter->cvR2, filter->cvR2, 60, UCHAR_MAX, cv::THRESH_BINARY); /* (R > 60) */
cv::threshold (filter->cvRp, filter->cvRp, 0.42, UCHAR_MAX, cv::THRESH_BINARY); /* (R'> 0.4) */
cv::threshold (filter->cvRp2, filter->cvRp2, 0.6, UCHAR_MAX, cv::THRESH_BINARY_INV); /* (R'< 0.6) */
cv::threshold (filter->cvGp, filter->cvGp, 0.28, UCHAR_MAX, cv::THRESH_BINARY); /* (G'> 0.28) */
cv::threshold (filter->cvGp2, filter->cvGp2, 0.4, UCHAR_MAX, cv::THRESH_BINARY_INV); /* (G'< 0.4) */
/* Combine all 3 thresholded color components, so that an output pixel will only
be white (255) if the H, S and V pixels were also white. */
2018-12-01 21:48:53 +00:00
cv::bitwise_and (filter->cvR2, filter->cvRp, filter->cvSkinPixels2);
cv::bitwise_and (filter->cvRp, filter->cvSkinPixels2,
filter->cvSkinPixels2);
cv::bitwise_and (filter->cvRp2, filter->cvSkinPixels2,
filter->cvSkinPixels2);
cv::bitwise_and (filter->cvGp, filter->cvSkinPixels2,
filter->cvSkinPixels2);
cv::bitwise_and (filter->cvGp2, filter->cvSkinPixels2,
filter->cvSkinPixels2);
filter->cvSkinPixels2.convertTo (filter->cvdraft, filter->cvdraft.type (),
1.0, 0.0);
cv::cvtColor (filter->cvdraft, filter->cvRGB, cv::COLOR_GRAY2RGB);
}
/* After this we have a RGB Black and white image with the skin, in
filter->cvRGB. We can postprocess by applying 1 erode-dilate and 1
dilate-erode, or alternatively 1 opening-closing all together, with
the goal of removing small (spurious) skin spots and creating large
connected areas */
if (filter->postprocess) {
2018-12-01 21:48:53 +00:00
cv::split (filter->cvRGB, channels);
filter->cvChA = channels.at (0);
2018-12-01 21:48:53 +00:00
cv::Mat element =
cv::getStructuringElement (cv::MORPH_RECT, cv::Size (3, 3),
cv::Point (1, 1));
cv::erode (filter->cvChA, filter->cvChA, element, cv::Point (1, 1), 1);
cv::dilate (filter->cvChA, filter->cvChA, element, cv::Point (1, 1), 2);
cv::erode (filter->cvChA, filter->cvChA, element, cv::Point (1, 1), 1);
2018-12-01 21:48:53 +00:00
cv::cvtColor (filter->cvChA, filter->cvRGB, cv::COLOR_GRAY2RGB);
}
2018-12-01 21:48:53 +00:00
filter->cvRGB.copyTo (outimg);
return GST_FLOW_OK;
}