Premium
On the training patterns of a neural network for target localization in the spatial domain
Author(s) -
Bermani E.,
Caorsi S.,
Massa A.,
Raffetto M.
Publication year - 2000
Publication title -
microwave and optical technology letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.304
H-Index - 76
eISSN - 1098-2760
pISSN - 0895-2477
DOI - 10.1002/1098-2760(20010205)28:3<207::aid-mop17>3.0.co;2-7
Subject(s) - sampling (signal processing) , lossy compression , artificial neural network , electric field , bandwidth (computing) , point (geometry) , microwave , domain (mathematical analysis) , computer science , space (punctuation) , mathematics , artificial intelligence , physics , computer vision , mathematical analysis , geometry , telecommunications , filter (signal processing) , quantum mechanics , operating system
In this letter, we show that some spatial sampling requirements for the scattered electric field are necessary in order to obtain good performances when a neural‐network approach is applied to the solution of target localization problems in the spatial domain. By means of some examples, concerning dielectric cylinders either in free space or buried in a lossy half space, we point out the dependence of the object localization on the spatial sampling rate of the scattered electric field and its relation to the field spatial bandwidth. © 2001 John Wiley & Sons, Inc. Microwave Opt Technol Lett 28: 207–209, 2001.