Fast Jitter Stabilization of Stationary High-Speed Video

Abstract

This report discusses a fast algorithm for image stabilization and jitter reduction for stationary cameras subjected to moderate shaking or vibrational forces. The localized corner-tracking algorithm was developed to stabilize high resolution, mid-wave infrared video but can be used for cameras in other spectrums, including the visible spectrum. The stabilization transform developed can be used to align the cameras field of view to a reference image frame so that measurements throughout a sequence of video are consistent. The algorithm is shown to reduce the effects of camera shake to a level comparable to state-of-the-art methods such as speeded-up robust feature (SURF) registration, while processing in a small fraction of the time (at least 60x faster), making it useful for high-speed video applications.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 2022
Accession Number
AD1187366

Entities

People

  • Ryan Decker
  • Steven Manole

Tags

Communities of Interest

  • Sensors
  • Weapons Technologies

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence
  • Cameras
  • Computations
  • Computer Science
  • Computer Vision
  • Detectors
  • Digital Images
  • Graphics Processing Unit
  • High Resolution
  • Image Processing
  • Image Registration
  • Measurement
  • Parallel Computing
  • Spectra
  • Stationary
  • Two Dimensional
  • Visible Spectra

Readers

  • Computer Vision.
  • Control Systems Engineering.
  • Image Processing and Computer Vision.