Optodigital implementation of a neural network using a joint transform correlator based in a Hopfield inner product model for character recognition

A. Serrano-Heredia; C. M. Hinojosa; R. Ponce; V. Arrizon

doi:10.1117/12.506399

23 October 2003 Optodigital implementation of a neural network using a joint transform correlator based in a Hopfield inner product model for character recognition

A. Serrano-Heredia, C. M. Hinojosa, R. Ponce, V. Arrizon

Author Affiliations +

Proceedings Volume 5202, Optical Information Systems; (2003) https://doi.org/10.1117/12.506399
Event: Optical Science and Technology, SPIE's 48th Annual Meeting, 2003, San Diego, California, United States

Abstract

Systems for automatic pattern recognition can be performed by Artificial Neural Networks and Optical Correlators. Here, we present the design and implementation of a scheme which takes the advantages of both systems to develop an hybrid opto-digital processor, with applications in character recognition. The implementation of this system is based in the Hopfield inner products model using a Joint Transform Correlator. The procedure of recognition has the following steps: since a correlation peak is proportional to the inner product, the Hopfield method computes the inner product of the input and each memory using the Hybrid Opto-Digital Joint Transform Correlator. The second step performs a multiplication between the inner product and its respective memory, all this scaled images are added to get the future state of the input. The associative memory is replaced by two images with information of all images in the memory, this memories are added in the last step. The signal output is threshold and feedback as an input for the next iteration. The process stops when the output image does not change in the next iteration. The final image corresponds to the closest image in the memory of the signal input. This implementation is strong and has low cost, with potential applications for real time pattern recognition.

Citation Download Citation

A. Serrano-Heredia, C. M. Hinojosa, R. Ponce, and V. Arrizon "Optodigital implementation of a neural network using a joint transform correlator based in a Hopfield inner product model for character recognition", Proc. SPIE 5202, Optical Information Systems, (23 October 2003); https://doi.org/10.1117/12.506399

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