A projection local image descriptor

  • Authors:
  • D. V. Sorokin;A. S. Krylov

  • Affiliations:
  • Centre for Biomedical Image Analysis, Faculty of Informatics, Masaryk University, Brno, Czech Republic;Laboratory of Mathematical Methods of Image Processing, Faculty of Computational Mathematics and Cybernetics, Moscow State University, Moscow, Russia

  • Venue:
  • Pattern Recognition and Image Analysis
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a projection local image descriptor. The projection local image descriptor construction is based on the local image expansion into the set of Gauss-Laguerre circular harmonic functions in the support region of a keypoint. The keypoints detection method is considered. It is also based on the analysis of image projections on the set of Gauss-Laguerre circular harmonic functions. The efficient technique for descriptors elements computation is introduced. The 2D Hermite projection method is used to accelerate the local image descriptors construction. The proposed approach is tested on the task of image points matching. The test results approved the ability of acceleration of descriptors construction process up to several times.