Wavelength calibration based on back propagation neural network

A method based on Back Propagation Neural Network (BPNN) for wavelength calibration of Spectrometer is proposed in this paper. An appropriate neural network is constructed to map pixel location in CCD to the wavelength with any given grating position. In this method, input vector of the network invo...

Full description

Saved in:
Bibliographic Details
Main Authors: Liang Zhang, Yinzhen Dai, Chun Lin, Ruiqi Lyu, Lei Wang, Tianlin Hu
Format: Conference Proceeding
Language:English
Subjects:
Calibration
Charge coupled devices
Equations
grating spectrometer
Gratings
neural network
Neural networks
Training
wavelength calibration
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A method based on Back Propagation Neural Network (BPNN) for wavelength calibration of Spectrometer is proposed in this paper. An appropriate neural network is constructed to map pixel location in CCD to the wavelength with any given grating position. In this method, input vector of the network involves the center wavelength of the spectrometer and the pixel location, output vector of the network is the spectral wavelength that has been corrected. And standard spectral lines from a Neon light are used for training of the neural network Compared with traditional method based on correction formula, this method reduces the mean error by 55.6% with the center wavelength range from 608nm to 668nm.
ISSN:2163-5048
2163-5056
DOI:10.1109/ICASID.2014.7064972