As for chromatic confocal sensor system with limited computational capacity, a fast peak extraction algorithm with considerate accuracy is in urgent demand. However, current peak extraction algorithms such as the centroid algorithm (CA) and nonlinear fitting algorithms can not balance the accuracy and computational efficiency. Thus, we propose an accurate peak extraction algorithm with good computational efficiency called corrected differential fitting algorithm (CDFA). At first, the differential signal derived from the original axial response signal is linearly fitted for initial peak extraction. Then corresponding systematic error of this linear fitting operation is analyzed using a first-order linear nonhomogeneous differential equation. At last, error compensation, that is, the solution to this equation is implemented with an introduction of "sum differences of sampling intensity". The performance of CDFA is compared with two conventional peak extraction algorithms including the CA and Gaussian fitting algorithm (GFA) using Monte Carlo simulations. CDFA is found to have a comparable accuracy performance with GFA while have a much higher computational efficiency.

Date Published: 7 March 2019
PDF: 9 pages
Proc. SPIE 11053, Tenth International Symposium on Precision Engineering Measurements and Instrumentation, 1105322 (7 March 2019); doi: 10.1117/12.2511734

Published in SPIE Proceedings Vol. 11053:
Tenth International Symposium on Precision Engineering Measurements and Instrumentation
Jiubin Tan; Jie Lin, Editor(s)