This article is cited in 11 papers

IMAGE PROCESSING, PATTERN RECOGNITION

Comparison of binary feature points descriptors of images under distortion conditions

E. A. Krasnabayeu^a, D. V. Chistobayeu^b, A. L. Malyshev<sup>a

a</sup> Vitebsk State University named after P.M. Masherov, Vitebsk, Belarus
^b Design Bureau “Display”, Vitebsk, Belarus

Abstract: The article is devoted to the review and analysis of binary descriptors of feature points of objects in digital images under distortion conditions. An overview of the BRIEF, ORB, BRISK, FREAK, AKAZE, LATCH methods is given. The evaluation of properties of the descriptors on sample images is performed. The paper addresses problems of using these methods for real time image processing.

Keywords: digital image processing, pattern recognition, image analysis, feature detection, feature description, feature matching.

Received: 19.01.2018
Accepted: 10.04.2019

DOI: 10.18287/2412-6179-2019-43-3-434-445