For the first time the possibility for parallel optical processing and recognition of interference patterns is shown. Sixty-four separated channels with multi-exposured holographic and low-cut filters are used The input interference patterns are calculated by finite element approach for penny-shaped crack in isotropic materials, surface defects and fault in laminate composite material. The set-up of an optical correlator for holographic recording of the filters and parallel processing of interference patterns is shown on Fig.1. High-resolution silver halide light sensitive material HP-650, made by the Bulgarian Academy of Sciences with different developers is used for holographic recording, creation of multiplicator and low-cut filters. Nine exposed holographic recordings were made for three different angles of rotation (-15 deg, 0, and +15 deg) and three different scales (-15%, 0, and +15%) of the interference patterns. Scale, shift and rotation invariant filtering has been obtained for a large angle of rotation of the input images The computer simulation, using a linear model of the system, is in good agreement with the experimental results. The differences could be explained by the nonlinearity of the multiexposure holographic recording For further development of the system and creation of the common filters for parallel optical processing of real two-dimensional interferometric data, a liquid crystal page composer and dynamic holographic recording in photorefractive materials are planed to be used.
[1]
B V Kumar,et al.
Multiple circular-harmonic-function correlation filter providing specified response to in-plane rotation.
,
1996,
Applied optics.
[2]
H H Arsenault,et al.
Scale-adapted wavelet transform.
,
1996,
Optics letters.
[3]
J. Álvarez-Borrego,et al.
Circular harmonic filters for the recognition of marine microorganisms.
,
1997,
Applied optics.
[4]
D. Casasent,et al.
Performance evaluation of minimum average correlation energy filters.
,
1991,
Applied optics.
[5]
Wolfgang Osten,et al.
Knowledge-assisted evaluation of fringe patterns for automatic fault detection
,
1994,
Optics & Photonics.
[6]
H H Arsenault,et al.
Object detection from a real scene using the correlation peak coordinates of multiple circular harmonic filters.
,
1989,
Applied optics.
[7]
B Javidi,et al.
Technique for reducing the redundant and self-correlation terms in joint transform correlators.
,
1993,
Applied optics.
[8]
E Dalsgaard,et al.
Synthetic circular-harmonic phase-only filter for shift, rotation, and scaling-invariant correlation.
,
1995,
Applied optics.
[9]
D Mendlovic,et al.
Shift and projection invariant pattern recognition using logarithmic harmonics.
,
1990,
Applied optics.
[10]
Jean-Pierre Huignard,et al.
Performance of a photorefractive joint transform correlator for fingerprint identification
,
1995
.