FEATURE DETECTORS IN THE VIOLA-JONES METHOD ON THE BASIS OF FINITE AUTOMATA

P. . Skorynin


Read the full article 

Abstract

Modification of the Viola-Jones method, one of the most effective methods for image classification, is
considered. Feature detectors built on the base of finite automata are used instead of the rectangular features detectors in the proposed modification. The use of finite automata allows for reducing the number of levels in the cascade and the number of detectors on each level due to the fact that the automata-based detector can detect more complicated features than simple rectangles.


Keywords: images classification, finite automata, machine learning
Copyright 2001-2017 ©
Scientific and Technical Journal
of Information Technologies, Mechanics and Optics.
All rights reserved.

Яндекс.Метрика