HYPERCOMPLEX ALGEBRAS AS UNIFIED LANGUAGE FOR IMAGE PROCESSING AND PATTERN RECOGNITION PART 1. CLIFFORDEAN MODELS OF MULTICHANNEL IMAGES: chapter in book

Valeriy G. Labunets, Juriy G. Smetanin, Victor P. Chasovskikh, Ekaterina Ostheimer

科研成果: Chapter同行评审

摘要

In this work, we assume that a brain in the visual cortex (VC) operates with Clifford numbers when it calculates hypercomplex-valued invariants of an image as it recognizes it. Clifford algebras generalize the algebras of complex numbers, quaternions and octonions. Of course, the algebraic nature of hypercomplex numbers must correspond to the spaces with respect to geometrically perceivable properties. For recognition of 2-D (bichromatic), 3-D (color), and n-D (multi-channel) images, we turn the perceptual spaces into corresponding Clifford algebras (and call them the VC-perceptual algebras). This approach gives full representation of how algebraic structures can possess image features and how algebraic structures can be used in different visual systems. It is our aim to show that the use of Clifford algebras fits more naturally to the tasks of recognition of multicolor patterns than does the use of color vector spaces. One can argue that nature has, through evolution, also learned to utilize properties of Cliffordean numbers.
源语言English
主期刊名ADVANCES IN INFORMATION TECHNOLOGIES, TELECOMMUNICATION, AND RADIOELECTRONICS
主期刊副标题сборник статей
编辑S. Kumkov, S. Shabunin, S. Singellakis
出版地点Cham
出版商Springer
3-19
页数17
ISBN(印刷版)978-3-030-37513-3
DOI
Published - 2020

出版系列

姓名Сер. Innovation and Discovery in Russian Science and Engineering (IDRSE)

GRNTI

  • 27.17.00

指纹

探究 'HYPERCOMPLEX ALGEBRAS AS UNIFIED LANGUAGE FOR IMAGE PROCESSING AND PATTERN RECOGNITION PART 1. CLIFFORDEAN MODELS OF MULTICHANNEL IMAGES: chapter in book' 的科研主题。它们共同构成独一无二的指纹。

引用此