rs-multicam sample

Overview

The multicam sample demonstrates the ability to use the SDK for streaming and rendering multiple devices simultaneously.

Expected Output

The application opens and renders a mosaic view of all automatically-selected streams provided by the connected devices. (The selection is device-tailored and takes into account USB-type constrains).
Each tile displays an unique stream produced by a specific camera. The stream name appear at the top left.

In the following snapshot we use five Intel® RealSense™ devices to produce the mosaic: D435i, D415, SR300 and two T265 tracking cameras. Those are responsible to generate:

• Three Depth streams (D435i, D415, SR300)
• Three Color streams (D435i, D415, SR300)
• Two Pose streams (T265x2)
• Four Fisheye streams (T265x2)
• Six IMU streams (D435i, T265x2)

Alltogether the mosaic comprise of 18 simultaneous live feeds:

Code Overview

As with any SDK application we include the Intel RealSense Cross Platform API:

#include <librealsense2/rs.hpp>     // Include RealSense Cross Platform API

In this example we will also use the auxiliary library of example.hpp:

#include "example.hpp"              // Include short list of convenience functions for rendering

examples.hpp lets us easily open a new window and prepare textures for rendering.

The first object we use is a window, that will be used to display the images from all the cameras.

// Create a simple OpenGL window for rendering:
window app(1280, 960, "CPP Multi-Camera Example");

The window class resides in example.hpp and lets us easily open a new window and prepare textures for rendering.

Next, we define the objects to be used in the example.

rs2::context ctx;    // Create librealsense context for managing devices

rs2::colorizer              colorizer;      // Utility class to convert depth data RGB

std::vector<rs2::pipeline>  pipelines;

The rs2::context encapsulates all of the devices and sensors, and provides some additional functionalities. We employ the rs2::colorizer to convert depth data to RGB format.
In the example we use multiple rs2::pipeline objects, each controlling a lifetime of a single HW device.

The example's flow starts with listing and activating all the connected Intel® RealSense™ devices:

// Start a streaming pipe per each connected device
for (auto&& dev : ctx.query_devices())
{
    rs2::pipeline pipe(ctx);
    rs2::config cfg;
    cfg.enable_device(dev.get_info(RS2_CAMERA_INFO_SERIAL_NUMBER));
    pipe.start(cfg);
    pipelines.emplace_back(pipe);
}

First, we allocate rs2::pipeline object per recognized device. Note that we share the rs2::context object between all rs2::pipeline instances.

rs2::pipeline pipe(ctx);

To map the specific device to the newly-allocated pipeline we define rs2::config object, and assign it with the device's serial number.
Then we request rs::pipeline to start streaming and produce frames.

pipe.start(cfg);

Since we do not specify explicit stream requests, each device is configured internally to run a set of predefined stream profiles recommended for that specific device.

After adding the device, we begin our main loop of the application:

while (app)

Every application cycle we traverse the registered devices and retrieve all the available frames:

// Collect the new frames from all the connected devices
std::vector<rs2::frame> new_frames;
for (auto &&pipe : pipelines)
{
   rs2::frameset fs;
   if (pipe.poll_for_frames(&fs))
   {
       for (rs2::frame& f : fs)
           new_frames.emplace_back(f);
   }
}

Each rs::pipeline produces a synchronized collection of frames for all streams configured for its allocated device. These are contained in rs2::frameset object.
The rs2::frameset itself is an wrapper for a composite_frame, which can holds more than a single type of frame.

To minimize UI impact we're using non-blocking frames polling method:

    if (pipe.poll_for_frames(&fs))

In order to simplify the presentation, we split those rs2::frameset containers into separate frames and store them with a standard C++ container for later use:

for (rs2::frame& f : fs)
    new_frames.emplace_back(f);

The Depth data is delivered as uint16_t type which cannot be rendered directly, therefore we use rs2::colorizer to convert the depth representation into human-readable RGB map:

// Convert the newly-arrived frames to render-friendly format
for (const auto& frame : new_frames)
{
    render_frames[frame.get_profile().unique_id()] = colorizer.process(frame);
}

And finally send the collected frames to update the openGl mosaic:

    app.show(render_frames);