
For FullText PDF, please login, if you are a member of IEICE,
or go to Pay Per View on menu list, if you are a nonmember of IEICE.

A Fast and Adaptive Imaging Algorithm for the Optical Array Imaging System
Osamu IKEDA
Publication
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Vol.E80A
No.6
pp.10921098 Publication Date: 1997/06/25
Online ISSN:
DOI:
Print ISSN: 09168508 Type of Manuscript: PAPER Category: Digital Signal Processing Keyword: digital signal processing, optimization, algorithm, adaptiveness, image processing,
Full Text: PDF(537.3KB)>>
Summary:
An optical array imaging system has been presented in previous articles. In this system, first, the object is illuminated with laser light sequentially from each of the array elements and the reflected field is detected as interferogram. The interferograms obtained are then spatially heterodynedetected on a computer to extract the signal components, that is, array data. Then, the eigenvector of the largest eigenvalue is derived by applying the power method to the array data and it is beamsteered to get images of the object. The algorithm gives good images for most objects, but it fails to work for some objects. It was shown that using a subset of the array data may solve the problem, but that finding the corresponding optimum subaperture is quite timeconsuming. In this paper, first, the integral equation describing the system is solved for a general class of object, to make clear the conditions for the eigenvector to form a sharp beam. Second, the imaging algorithm is sped up to a great degree by optimizing only the illuminating aperture in a coarse fashion. Third, the rate of convergence of the power method is adaptively estimated in the algorithm to make the eigenvector derivation reliable. Finally the improved algorithm is investigated using both computergenerated and experimentally obtained array data.

