Neural learning algorithm for halftoning

Most processes used for halftoning consist of linear and nonlinear elements. Neural networks offer the possibility of combining these elements in a general and flexible structure. Image binarization methods can be analysed and transfered to neural structures and typical neural learning algorithms of...

Full description

Saved in:
Bibliographic Details
Published inOptics communications Vol. 113; no. 4; pp. 360 - 364
Main Authors Tuttaß, T., Bryngdahl, O.
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 1995
Elsevier Science
Subjects
Applied sciences
Artificial intelligence
Computer science; control theory; systems
Electric, optical and optoelectronic circuits
Electronics
Exact sciences and technology
Image processing
Information, signal and communications theory
Neural networks
Pattern recognition. Digital image processing. Computational geometry
Signal processing
Telecommunications and information theory
Fourier transformation
Function block diagram
Grey level image
Image processing
Iterative process
Algorithmics
Theoretical study
Neural network
Learning algorithm
Half tone image
Adaptation
Mean square error
Online AccessGet full text
ISSN0030-4018
1873-0310
DOI10.1016/0030-4018(94)00570-K

Cover

More Information
Summary:Most processes used for halftoning consist of linear and nonlinear elements. Neural networks offer the possibility of combining these elements in a general and flexible structure. Image binarization methods can be analysed and transfered to neural structures and typical neural learning algorithms offer new ways to treat the halftoning problem. We examine a simple learning algorithm and demonstrate the difficulties and possibilities concerning the halftoning problem.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0030-4018
1873-0310
DOI:10.1016/0030-4018(94)00570-K