dlib/examples/ffmpeg_rtsp_ex.cpp

149 lines
4.8 KiB
C++
Raw Normal View History

2023-04-09 22:34:31 +08:00
// The contents of this file are in the public domain. See LICENSE_FOR_EXAMPLE_PROGRAMS.txt
/*
This is an example illustrating the use of the ffmpeg wrappers.
It decodes a video, pushes frames over an RTSP client connection and receives on another RTSP server connection.
Please see all the other ffmpeg examples:
- ffmpeg_info_ex.cpp
- ffmpeg_video_decoding_ex.cpp
[FFmpeg] decoding and demuxing improvements (#2784) * typo * - added compile time information to audio object. Not convinced this is needed actually. I'm perfectly happy just using the ffmpeg::frame object. I'm pretty sure I'm the only user who cares about audio. - created resizing_args and resampling_args * smaller videos for unit tests * shorter videos for unit tests * - decoder and demuxer: you now resize or resample at the time of read. therefore you don't set resizing or resampling parameters in constructor, but you pass them to read() - added templated read() function - simplified load_frame() * inherit from resizing_args and resampling_args * reorganised the tests to segragate decoding, demuxing, encoding and muxing as much as possible * much more basic example * demxing examples split * examples * fixing examples * wip * Fix load_frame() * added frame - specific tests * - makes sense to have a set_params() method rather than constructing a new object and moving. I mean, it works and it absolutely does the right thing, and in fact the same thing as calling set_params() now, but it can look a bit weird. * notes on defaults and good pairings * Update ffmpeg_demuxer.h Watch out for `DLIB_ASSERT` statements. Maybe one of the unit tests should build with asserts enabled. * Update ffmpeg_details.h * Update ffmpeg_muxer.h * WIP * WIP * - simplified details::resizer - added frame::set_params() - added frame::clear() - forward packet directly into correct queue * pick best codec if not specified * added image data * warn when we're choosing an appropriate codec * test load_frame() * - for some reason, you sometimes get warning messages about too many b-frames. Resetting pict_type suppresses this. - you can move freshly decoded frames directly out. * callback passed to push() * I think it's prettier this way * WIP * full callback API for decoder * updated tests * updated example * check the template parameter is callable and has 1 argument first before getting it's first argument * Potential bug fix * - write out the enable_if's explictly. It's fine. I think it's clear what's going on if someone cares - guard push() with a boolean which asserts when recursion is detected * pre-conditions on callbacks: no recursion --------- Co-authored-by: pf <pf@me> Co-authored-by: Your name <you@example.com>
2023-05-16 09:24:47 +08:00
- ffmpeg_video_decoding2_ex.cpp
2023-04-09 22:34:31 +08:00
- ffmpeg_video_demuxing_ex.cpp
[FFmpeg] decoding and demuxing improvements (#2784) * typo * - added compile time information to audio object. Not convinced this is needed actually. I'm perfectly happy just using the ffmpeg::frame object. I'm pretty sure I'm the only user who cares about audio. - created resizing_args and resampling_args * smaller videos for unit tests * shorter videos for unit tests * - decoder and demuxer: you now resize or resample at the time of read. therefore you don't set resizing or resampling parameters in constructor, but you pass them to read() - added templated read() function - simplified load_frame() * inherit from resizing_args and resampling_args * reorganised the tests to segragate decoding, demuxing, encoding and muxing as much as possible * much more basic example * demxing examples split * examples * fixing examples * wip * Fix load_frame() * added frame - specific tests * - makes sense to have a set_params() method rather than constructing a new object and moving. I mean, it works and it absolutely does the right thing, and in fact the same thing as calling set_params() now, but it can look a bit weird. * notes on defaults and good pairings * Update ffmpeg_demuxer.h Watch out for `DLIB_ASSERT` statements. Maybe one of the unit tests should build with asserts enabled. * Update ffmpeg_details.h * Update ffmpeg_muxer.h * WIP * WIP * - simplified details::resizer - added frame::set_params() - added frame::clear() - forward packet directly into correct queue * pick best codec if not specified * added image data * warn when we're choosing an appropriate codec * test load_frame() * - for some reason, you sometimes get warning messages about too many b-frames. Resetting pict_type suppresses this. - you can move freshly decoded frames directly out. * callback passed to push() * I think it's prettier this way * WIP * full callback API for decoder * updated tests * updated example * check the template parameter is callable and has 1 argument first before getting it's first argument * Potential bug fix * - write out the enable_if's explictly. It's fine. I think it's clear what's going on if someone cares - guard push() with a boolean which asserts when recursion is detected * pre-conditions on callbacks: no recursion --------- Co-authored-by: pf <pf@me> Co-authored-by: Your name <you@example.com>
2023-05-16 09:24:47 +08:00
- ffmpeg_video_demuxing2_ex.cpp
2023-04-09 22:34:31 +08:00
- ffmpeg_video_encoding_ex.cpp
- ffmpeg_video_muxing_ex.cpp
*/
#include <cstdio>
#include <dlib/media.h>
#include <dlib/cmd_line_parser.h>
#include <dlib/gui_widgets.h>
using namespace std;
using namespace std::chrono;
using namespace std::chrono_literals;
using namespace dlib;
using namespace dlib::ffmpeg;
int main(const int argc, const char** argv)
try
{
command_line_parser parser;
parser.add_option("i", "input video", 1);
parser.add_option("codec", "video codec name. e.g. `h264`. Defaults to `mpeg4`", 1);
parser.add_option("height", "height of encoded stream. Defaults to whatever is in the video file", 1);
parser.add_option("width", "width of encoded stream. Defaults to whatever is in the video file", 1);
parser.set_group_name("Help Options");
parser.add_option("h", "alias of --help");
parser.add_option("help", "display this message and exit");
parser.parse(argc, argv);
const char* one_time_opts[] = {"i", "codec", "height", "width"};
parser.check_one_time_options(one_time_opts);
if (parser.option("h") || parser.option("help"))
{
parser.print_options();
return 0;
}
const std::string input_filepath = parser.option("i").argument();
// First, we open a video which we use to transmit and receive images over RTSP.
demuxer cap({input_filepath, video_enabled, audio_disabled});
if (!cap.is_open())
{
cout << "Failed to open " << input_filepath << endl;
return EXIT_FAILURE;
}
if (!cap.video_enabled())
{
cout << "This video does not contain images." << endl;
return EXIT_FAILURE;
}
const std::string url = "rtsp://0.0.0.0:8000/stream";
// We start 2 threads:
// - 1 for an RTSP server that listens for an incoming RTSP client and decodes the frames
// - 1 for an RTSP client that connects and pushes/muxes image frames.
std::thread rx{[&]
{
// This is an example that show-cases the usage of demuxer::args::format_options.
// This is used for AVFormatContext and demuxer-private options specific to the container.
// {"rtsp_flags", "listen"} tells the RTSP demuxer that it is a server
// {"rtsp_transport", "tcp"} configures RTSP over TCP transport. This way we don't loose any packets.
// We set a listening timeout of 5s.
demuxer cap([&] {
demuxer::args args;
args.filepath = url;
args.format_options["rtsp_flags"] = "listen";
args.format_options["rtsp_transport"] = "tcp";
args.connect_timeout = 5s;
return args;
}());
if (!cap.is_open())
{
cout << "Failed to receive connection from RTSP client" << endl;
return;
}
image_window win;
array2d<rgb_pixel> img;
[FFmpeg] decoding and demuxing improvements (#2784) * typo * - added compile time information to audio object. Not convinced this is needed actually. I'm perfectly happy just using the ffmpeg::frame object. I'm pretty sure I'm the only user who cares about audio. - created resizing_args and resampling_args * smaller videos for unit tests * shorter videos for unit tests * - decoder and demuxer: you now resize or resample at the time of read. therefore you don't set resizing or resampling parameters in constructor, but you pass them to read() - added templated read() function - simplified load_frame() * inherit from resizing_args and resampling_args * reorganised the tests to segragate decoding, demuxing, encoding and muxing as much as possible * much more basic example * demxing examples split * examples * fixing examples * wip * Fix load_frame() * added frame - specific tests * - makes sense to have a set_params() method rather than constructing a new object and moving. I mean, it works and it absolutely does the right thing, and in fact the same thing as calling set_params() now, but it can look a bit weird. * notes on defaults and good pairings * Update ffmpeg_demuxer.h Watch out for `DLIB_ASSERT` statements. Maybe one of the unit tests should build with asserts enabled. * Update ffmpeg_details.h * Update ffmpeg_muxer.h * WIP * WIP * - simplified details::resizer - added frame::set_params() - added frame::clear() - forward packet directly into correct queue * pick best codec if not specified * added image data * warn when we're choosing an appropriate codec * test load_frame() * - for some reason, you sometimes get warning messages about too many b-frames. Resetting pict_type suppresses this. - you can move freshly decoded frames directly out. * callback passed to push() * I think it's prettier this way * WIP * full callback API for decoder * updated tests * updated example * check the template parameter is callable and has 1 argument first before getting it's first argument * Potential bug fix * - write out the enable_if's explictly. It's fine. I think it's clear what's going on if someone cares - guard push() with a boolean which asserts when recursion is detected * pre-conditions on callbacks: no recursion --------- Co-authored-by: pf <pf@me> Co-authored-by: Your name <you@example.com>
2023-05-16 09:24:47 +08:00
while (cap.read(img))
2023-04-09 22:34:31 +08:00
win.set_image(img);
}};
std::this_thread::sleep_for(1s);
std::thread tx{[&]
{
// The muxer acts as an RTSP client, so we don't use {"rtsp_flags", "listen"}
// When using RTSP, it is usually a good idea to specify muxer::args::output_format = "rtsp"
2023-06-12 08:21:05 +08:00
// even though the URL has rtsp:// in its address. Whether or not you need to specify args.output_format = "rtsp"
2023-04-09 22:34:31 +08:00
// depends on your version of ffmpeg.
muxer writer([&] {
muxer::args args;
args.filepath = url;
args.output_format = "rtsp";
args.enable_image = true;
args.enable_audio = false;
args.format_options["rtsp_transport"] = "tcp";
args.args_image.codec_name = get_option(parser, "codec", "mpeg4");
args.args_image.h = get_option(parser, "height", cap.height());
args.args_image.w = get_option(parser, "width", cap.width());
args.args_image.fmt = cap.pixel_fmt();
args.args_image.framerate = cap.fps();
return args;
}());
if (!writer.is_open())
{
cout << "Failed to open rtsp server" << endl;
return;
}
frame f;
while (cap.read(f))
writer.push(std::move(f));
}};
tx.join();
rx.join();
return EXIT_SUCCESS;
}
catch (const std::exception& e)
{
printf("%s\n", e.what());
return EXIT_FAILURE;
}