VersaVIS

VersaVIS provides precise time synchronization for multi-camera and IMU systems to enable robust visual-inertial pose estimation and sensor fusion for mobile robotics and SLAM applications.


Key Features:

  • Time Synchronization: Achieves precise time synchronization between multiple cameras and an IMU to support accurate sensor fusion.
  • Exposure Compensation: Compensates for varying exposure times to maintain consistent data capture across different lighting conditions.
  • Host Clock Translation: Performs host clock translation to reconcile different system clocks and reduce synchronization errors from multiple clock sources.
  • Independent and Stereo Camera Triggering: Supports independent and stereo camera triggering configurations for flexible acquisition schemes.
  • Hardware, Firmware, and Software Integration: Integrates hardware, firmware, and software components to facilitate robust and accurate pose estimation.
  • Robust Delay Handling: Addresses internal sensor filtering, unpredictable operating system scheduling delays, and data transfer lags to improve temporal alignment.
  • Device Support: Supports a wide range of cameras and IMUs for adaptable sensor configurations.
  • Sub-millisecond Synchronization Accuracy: Demonstrates synchronization timing precision of less than 1 millisecond across multiple experiments.

Scientific Applications:

  • Visual-Inertial SLAM: Integrates visual data with inertial measurements to enhance the robustness and accuracy of SLAM systems.
  • Multi-Camera Systems: Enables synchronized acquisition for multi-camera mapping and reconstruction tasks.
  • Multi-Modal Mapping: Supports integration of various sensor modalities for combined mapping applications.
  • Reconstruction and Object-Based Mapping: Aids in detailed environmental reconstruction and object-based mapping.

Methodology:

Handles sensor synchronization by addressing internal sensor filtering, unpredictable operating system scheduling delays, and data transfer lags.

Topics

Details

Programming Languages:
C++, MATLAB, Python
Added:
1/18/2021
Last Updated:
3/12/2021

Operations

Publications

Tschopp F, Riner M, Fehr M, Bernreiter L, Furrer F, Novkovic T, Pfrunder A, Cadena C, Siegwart R, Nieto J. VersaVIS—An Open Versatile Multi-Camera Visual-Inertial Sensor Suite. Sensors. 2020;20(5):1439. doi:10.3390/s20051439. PMID:32155749. PMCID:PMC7085520.

PMID: 32155749
PMCID: PMC7085520
Funding: - National Centre of Competence in Research Robotics: P3Y9P5 - Siemens Mobility GmbH, Germany: none