Projekt: HECTOR (Hand Eye Calibration Toolkit for One-arm Robotsystems)
Forscher: Jordt A. , Siebel N.
HECTOR (Hand Eye Calibration Toolkit for One-arm Robot systems) is an application to support the development and deployment of calibration algorithms. It has a wide range of features:- Real time 3D visualisation of arbitrary objects
- Integrated robot simulation, including 3D visualisation, inverse kinematics and eye-in-hand camera simulation
- Remote control of arbitrary robot systems
- Integration of real time camera data, including image processing and visualisation features
- Sync mode, to compare simulated and real camera robot system data or to visualise motion planing of the robot in real time.
- An interactive console enables the user to directly influence the running tests. It features auto completion and history. The console is able to run scripts, so large test series can be automated.
It is easy to extend the capabilities of HECTOR by writing new console commands, including own C++ source code into the existing framework. To make this easy, the core system of HECTOR provides a versatile callback structure. Functions can be registered to be called at certain events (arrival of the robot at a new destination, new image data available, idle time, etc.), and custom event triggers can be easily defined and triggered.HECTOR is based on the Perception Action Components Library (PACLib) by the Cognitive Systems group. Images received from the HECTOR API are in the PACImage format, providing a wide range of complex image processing components.