Preprocessing and classification of objects based on neural network models built using memristors' elements

Evgenii Semenishchev; Sergey Zhukov; Viacheslav Voronin; Andrey Gribkov; Aleksandr Zelensky

doi:10.1117/12.2691430

27 November 2023 Preprocessing and classification of objects based on neural network models built using memristors' elements

Evgenii Semenishchev, Sergey Zhukov, Viacheslav Voronin, Andrey Gribkov, Aleksandr Zelensky

Author Affiliations +

Proceedings Volume 12769, Optical Metrology and Inspection for Industrial Applications X; 127691O (2023) https://doi.org/10.1117/12.2691430
Event: SPIE/COS Photonics Asia, 2023, Beijing, China

Conference Poster

Abstract

The article proposes the use of three directions of development at once to solve the problem of dividing objects into classes. The first direction uses the formation of a computationally simple multi-criteria method for smoothing data in windows of complex adaptive forms, adapting the method of dividing objects into background/structure, and simplifying images. The approaches being developed are intended for implementation on low-computing devices with the ability to parallelize processes. The second direction being worked on is the formation of a model of the structure of a neuron organized on the basis of the use of memristor structures. The paper presents an approach to the formation of such structures, provides the characteristics of such devices, and describes methods for combining analog and digital parts to implement memory or control systems. The final direction discussed in the article is the formation of a neural network for the classification of simple objects based on a model of new neurons and data preprocessing. To test the approach proposed in the work, studies were carried out on a set of test data obtained by a sensor (simple sensor) system. The generated data array for evaluating efficiency is limited by a time window and has real noise (errors). The work provides assessments of effectiveness, recommendations for the selection of parameters and presents requirements for the type and form of the analyzed data.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Evgenii Semenishchev, Sergey Zhukov, Viacheslav Voronin, Andrey Gribkov, and Aleksandr Zelensky "Preprocessing and classification of objects based on neural network models built using memristors' elements", Proc. SPIE 12769, Optical Metrology and Inspection for Industrial Applications X, 127691O (27 November 2023); https://doi.org/10.1117/12.2691430

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available