Integrated optical processors in pattern recognition

Abstract Integrated coherent optical processors are presented and their applications to the field of pattern recognition are considered. It is noted that having coherent optical processors in integrated forms removes many constraints on field applications of conventional optical processors. However, integration introduces some new problems that could degrade optical processor performance. The effects of first-order sources of degradation on the optical architectures proposed here are investigated. How they would affect the processing algorithms developed here is also reported. The problem of preprocessing remotely sensed data to achieve high data rate reduction is addressed. Two distance measure optical processors that operate on the principle of interferometry are proposed. Processing algorithms (parametric and nonparametric) that take full advantage of the statistical dependence of the sensed data on their neighbors as well as the high processing speed of the optical processors are developed. System performance is evaluated and it is shown that intensity fluctuations of the optical source (laser) can have an adverse effect on the entire system.

[1]  David A. Landgrebe,et al.  The development of a spectral-spatial classifier for earth observational data , 1980, Pattern Recognit..

[2]  R. A. Holmes,et al.  The physical basis of system design for remote sensing in agriculture , 1969 .

[3]  K.-s. Fu.  Pattern Recognition in Remote Sensing of the Earth's Resources , 1976, IEEE Transactions on Geoscience Electronics.

[4]  R. J. Ginther,et al.  Optical waveguides formed by thermal migration of ions in glass. , 1973, Applied optics.

[5]  J. Armstrong,et al.  VI Experimental Studies of Intensity Fluctuations in Lasers , 1967 .

[6]  A. B. Vander Lugt,et al.  Signal detection by complex spatial filtering , 1964, IEEE Trans. Inf. Theory.

[7]  Joseph T. Boyd,et al.  Effect of waveguide optical scattering on the integrated optical spectrum analyzer dynamic range , 1978 .

[8]  Kazuhiro Daikoku,et al.  A proposal on optical fibre transmission systems in a low-loss 1.0–1.4 μm wavelength region , 1977 .

[9]  K. M. van Vliet,et al.  Noise Limitations in Solid State Photodetectors , 1967 .

[10]  R. Kettig,et al.  Classification of Multispectral Image Data by Extraction and Classification of Homogeneous Objects , 1976, IEEE Transactions on Geoscience Electronics.

[11]  Oscar Weinstein,et al.  Design Challenges of the Thematic Mapper , 1980, IEEE Transactions on Geoscience and Remote Sensing.

[12]  C. M. Verber,et al.  Feasibility investigation of integrated optics Fourier transform devices. [holographic subtraction for multichannel data preprocessing] , 1977 .

[13]  I. Deyhimy,et al.  High sensitivity optical receivers for 1.0-1.4 µm fiber-optic systems , 1978 .

[14]  P. M. Narendra,et al.  A non-parametric clustering scheme for landsat , 1977, Pattern Recognit..

[15]  B.J. Thompson.  Hybrid processing systems—An assessment , 1977, Proceedings of the IEEE.

[16]  David Casasent.  A Hybrid Digital/Optical Computer System , 1973, IEEE Transactions on Computers.

[17]  David Casasent.  Coherent Optical Computing , 1979, Computer.

[18]  D. A. Landgrebe.  Machine Processing for Remotely Acquired Data , 1973 .