As a kind of film device, band-pass filter is widely used in pattern recognition, infrared detection, optical fiber communication, etc. In this paper, an algorithm for automatic measurement of band-pass filter quality criterion is proposed based on the proven theory calculation of derivate spectral transmittance of filter formula. Firstly, wavelet transform to reduce spectrum data noises is used. Secondly, combining with the Gaussian curve fitting and least squares method, the algorithm fits spectrum curve and searches the peak. Finally, some parameters for judging band-pass filter quality are figure out. Based on the algorithm, a pipeline for band-pass filters automatic measurement system has been designed that can scan the filter array automatically and display spectral transmittance of each filter. At the same time, the system compares the measuring result with the user defined standards to determine if the filter is qualified or not. The qualified product will be market with green color, and the unqualified product will be marked with red color. With the experiments verification, the automatic measurement system basically realized comprehensive, accurate and rapid measurement of band-pass filter quality and achieved the expected results.
[1]
William H. Press,et al.
Numerical recipes: the art of scientific computing, 3rd Edition
,
2007
.
[2]
R. J. Peppin.
An Introduction to Random Vibrations, Spectral and Wavelet Analysis
,
1994
.
[3]
Sha Ji-chang.
Research of Piecewise Cubic Curve-fitting Method Based on Least-square Principle
,
2007
.
[4]
Xin-She Yang,et al.
Introduction to Algorithms
,
2021,
Nature-Inspired Optimization Algorithms.
[5]
Hongwei Guo,et al.
A Simple Algorithm for Fitting a Gaussian Function
,
2012
.