Overview

Intel® Robot DevKit (RDK) Project contains robotics related open source software components under ROS2 framework for RealSense based perceptual computation, neuron network based object and people face detection, object tracking and 3D localization, SLAM, navigation, visual manipulation for industry robot, and a bunch of tools for development and debugging.

Key Packages

ros2_intel_realsense

ros2_intel_realsense for using Intel® RealSense™ cameras (D400 series and T265) with ROS2.

ros2_object_analytics

ros2_object_analytics is a group of ROS2 packages for real-time object detection, localization and tracking. These packages aim to provide real-time object analyses over RGB-D camera inputs, enabling ROS developer to easily create amazing robotics advanced features, like intelligent collision avoidance and semantic SLAM.

ros2_openvino_toolkit

The OpenVINO™ toolkit quickly deploys applications and solutions that emulate human vision. Based on Convolutional Neural Networks (CNN), the toolkit extends computer vision (CV) workloads across Intel® hardware, maximizing performance.

This project is a ROS2 wrapper for CV API of OpenVINO™, providing the following features:

  • Support CPU and GPU platforms
  • Support standard USB camera and Intel® RealSense™ camera
  • Support Video or Image file as detection source
  • Face detection:
    • Emotion recognition
    • Age and gender recognition
    • Head pose recognition
  • Object detection
  • Object segmentation
  • Demo application to show above detection and recognitions

navigation2

The ROS 2 Navigation System is the control system that enables a robot to autonomously reach a goal state, such as a specific position and orientation relative to a specific map. Given a current pose, a map, and a goal, such as a destination pose, the navigation system generates a plan to reach the goal, and outputs commands to autonomously drive the robot, respecting any safety constraints and avoiding obstacles encountered along the way.

ros2_grasp_library

ROS2 Grasp Library enables state-of-the-art CNN based deep learning grasp detection algorithms on ROS2 for visual based industrial robot manipulation.

tools

To facilitate the development work, several tools are provided, including rtmonitor to monitor and capture real time performance metrics of ROS2 C++ Application, and rdk_perf to measure ROS2 topics and show result in rqt at run time.