Scalable Hardware and Software Procedures for N-ocular Event-Based Stereo System Calibration

Abstract

Pre-calibrated stereo-based vision approaches offer the capability to resolve 3D feature information instantaneously without the need to register image information and compute camera geometric transformations. Moreover, event-based cameras (EBCs) promise to enhance this capability due to their inherent compression of scene information and high temporal resolution. However, this ability requires careful calibration of the stereo system. Due to their recent inception, current approaches to enable this capability for EBCs are not disseminated amongst the broad community. This report presents a set of hardware and software techniques that can be used and scaled to calibrate N-ocular wide-baseline stereo systems of EBCs. The design of two hardware instruments and two software algorithms are detailed. These approaches are ultimately applied to extract the intrinsic and extrinsic parameters of a four-camera EBC system with millimeter resolution.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 19, 2024
Accession Number
AD1219229

Entities

People

  • Ethan K Eldridge
  • Jonah P. Sengupta
  • Matthew R Devine

Organizations

  • United States Army Research Laboratory

Tags

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Fluid Dynamics.
  • Image Processing and Computer Vision.