Use this URL to cite or link to this record in EThOS:
Title: Frequency-division-multiplexing technique for imaging metrology
Author: Bledowski, Ian A.
ISNI:       0000 0004 5346 1329
Awarding Body: Cranfield University
Current Institution: Cranfield University
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Access from Institution:
An algorithm to multiplex multiple image captures simultaneously onto a single image sensor at full frame resolution was developed for imaging metrology. Parseval’s theorem was used to obtain the image intensity from image time-series of around typically 256 frames captured by the imaging sensor at typically 60 fps, though kHz frame rates are possible, hardware permitting. The time-series contained contributions from each image channel in the system, which were created by periodically modulating the intensity of the light source which defined that channel. The modulating time-series was converted to a frequency representation by Fourier transform and from that the channels could be identified by their peaks in the spectrum. Peaks corresponding to each channel were then isolated with a window function and Parseval’s theorem applied on a pixel by pixel basis to convert the signal strength back to an image containing the information from that channel only. The FDM algorithm was then applied to two imaging metrology methods. First, an in-plane, two-channel shearography system was multiplexed with FDM in such a way as to allow time-division multiplexed measurements to be taken on the same deformations with the same instrument so as to allow comparison of results from other methods. FDM was found to produce good quality results comparable with current methods. Interferometric planar Doppler velocimetry was performed, multiplexing the reference phase channel signal and a signal channel for both a wheel and a gas jet. FDM was found to suppress the effects of phase drifts in the system which would lead to velocity offsets in the results, and gave velocities which varied from the model by only up to ~5%. Finally, an error analysis was performed on the FDM algorithm, comparing the technique with time-averaging and single image capture through simulation and practical methods. It was shown that FDM strongly suppresses the noise and background in a measurement, and can produce good images from low intensity signals. It could be concluded that the FDM algorithm offers significant advantages over time-averaging a signal when applied to a multi-channel imaging metrology system.
Supervisor: Tatam, Ralph P.; James, Stephen W. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available