I am often asked to comment on the suitability of various topics and kinds of manuscripts for submission to Optical/Engineering, and one of the most frequently asked questions is related to the distinction between experimental and theoretical papers; i.e., should Optical Engineering contain papers of a theoretical nature? Another question I have encountered more than once has to do with the appropriateness of papers on image processing. Some individuals with whom I have spoken seem to have very definite ideas about these matters, but I find it difficult to be as confident as they.
The aim of this special issue of Optical Engineering on Optical Processing for the Space Station is to provide a forum for the presentation and future discussion of the full range of research and development activities in optics that have the potential for increasing the space-oriented mission automation capabilities during the Space Station era. Each paper is devoted to a specific technology and/or technique that is essential to the development and evolution of highly efficient spaceborne optical processors. A discussion is given on the potential use of artificial intelligence technologies such as expert systems to greatly enhance the overall system performance and capability of optical processors.
Opt. Eng. 25(7), 257808 (1 July 1986) doi:10.1117/12.7973912
TOPICS: Optical signal processing, Astronomical imaging, Control systems, Optical pattern recognition, Process control, Classification systems, Scene classification, Data processing, Optical correlators, Analog electronics
Optical information processing research aimed at Space Station automation applications is reviewed. The emphasis of the NASA Ames Research Center program is on intelligent optical pattern recognition and optical control processing. Attention is given to the primary functions of an overall scene understanding system: distortion-invariant optical feature generation, dimensionality reduction, object classification, and contextual information processing. A method of using synthetic discriminant functions to facilitate learning in a high-speed optical correlator is described. A discussion is presented of candidate analog and digital architectures for the optical implementation of state-estimation algorithms needed for the control of high-dimension dynamic systems. The multivariate system chosen for the optical control technology demonstration--a segmented, adaptive mirror and interferometrically based wavefront sensor--is also described.
Automated rendezvous, stationkeeping, and docking, as well as robot vison for space construction, require a compact, rugged system to provide two-and three-dimensional (range) images. The system proposed in this paper consists of a module having a transmitter array of laser diodes and a single detector/receiver that covers the full field of view (FOV) of the module. Each diode in the array is imaged on a pixel in the FOV. Sinusoidal intensity modulation is used such that the range at each pixel can be determined by phase comparison of the transmitted and received signals. The system FOV can be expanded by using multiple modules. The laser diode modulation frequency is initially low to determine coarse range without ambiguities. It is then increased in sequential steps to reach the desired accuracy for a given range. Because the scanning is accomplished by individually illuminating elements in the diode array, the scanning is achieved with no moving parts, and the pixels in the FOV can be randomly accessed. Preliminary results demonstrated a 15 cm range accuracy with an uncooperative target at a range of 13 m in bright sunlight.
Using space-variant pattern recognition of up to 256 3 x 3 patterns of l's and 0's in parallel and inserting image information sequentially in a well-defined pattern, we can construct an optical systolic cellular array processor for 3 x 3 neighborhoods that produces output points at one-third the rate at which points are input. This allows reprogrammable preprocessing of data input.
This paper introduces a novel approach to building a 2-D scanning optical correlator with large memory capacity for multiple-object tracking, with possible applications for telerobotic vision in the Space Station. The system consists of a tunable dye laser and a dispersive optical element such as a diffraction grating. The 2-D scanning operation is achieved by sequential tuning of the dye laser and rotation of the diffraction grating about the optical axis. The system capacity is greatly incresed by this wavelength-angle multiplexing scheme. In synthesizing the spatially multiplexed matched spatial filters, a monochromatic laser is used with a spatial carrier and scale compensation technique. Up to 1000 filters can be fabricated on a single holographic plate with a modest requirement of system space-bandwidth product. Several illustrating experimental demonstrations are presented.
Application of the Foley-Sammon transform to image classification is discussed. The computing procedures of the transform are presented for the case in which the total number of training images (M) is smaller than the dimension of the images to be classified (N). A set of computer simulation results shows that the Foley-Sammon transform provides good classification performance and that a proper number of training images is essential to ensure a low error rate of classification.
A compact color schlieren system designed for field measurement of materials processing parameters has been built and tested in a microgravity environment. Improvements in the color filter design and a compact optical arrangement allowed the system described here to retain the traditional advantages of schlieren, such as simplicity, sensitivity, and ease of data interpretation. Testing was accomplished by successfully flying the instrument on a series of parabolic trajectories on the NASA KC-135 microgravity simulation aircraft. A variety of samples of interest in materials processing were examined. Although the present system was designed for aircraft use, the technique is well suited to space flight experimentation. A major goal of this effort was to accommodate the main optical system within a volume approximately equal to that of a Space Shuttle middeck locker. Future plans include the development of an automated space-qualified facility for use on the Shuttle and Space Station.
A microcomputer-based real-time programmable optical signal processing system utilizing a magneto-optic spatial light modulator (MOSLM) and a liquid crystal light valve (LCLV) is described. This system can perform a myriad of complicated optical operations, such as image correlation, image subtraction, and matrix multiplication. Its important assets are its programmability and the capability of real-time addressing. These are important for telerobotic vision in space automation applications. Design specifications and suggestions for practical implementation of the system are discussed. Some preliminary experimental demonstrations are conducted to demonstrate applications of the proposed system to image correlation for optical pattern recognition, image subtraction for IC chip inspection, and matrix multiplication for optical computing.
A real-time optical associative retrieval technique is presented. The associative retrieval model enables a large amount of data to be stored and recalled by partial information optically in real time. The real-time capability is achieved by using an electronically addressed spatial light modulator based on the pocket-size liquid crystal display television. The potential application of the technique to the perceptive vision requirements in telerobotics for achieving NASA's goals of automation in space is described.
In many applications optical wavefields with reduced states of coherence are preferable to fully coherent fields. Gaussian Schell-model beams can be viewed as generalizations of the ordinary fundamental Gaussian laser beam by allowing a Gaussian variation of the transverse spatial coherence. Suitable propagation parameters for Gaussian Schell-model beams are stated, and their evolution in free space and through optical systems is analyzed. An illustrative geometrical approach, analogous to the propagation-circle method for laser beams, is also presented. Finally, the effects of spatial coherence on beam focusing are investigated, and applications of the algebraic and graphical methods to some particular problems are briefly discussed.
An object-localized reference mirror is used for the first time in an electronic speckle pattern interferometer (ESPI). This approach, together with the established technique of stroboscopic ESPI, is used to study the dynamic volume viscoelasticity of a submerged rubber sample undergoing sinusoidal vibration. These two modulation techniques allow manipulation of the fringe field describing the sample movement, helping to simplify data extraction.
The third- and fifth-order wavelength-shift aberrations for an off-axis holographic zone plate for the higher orders are discussed, and numerical calculations of the aberrations are presented. A stigmatic point image can be obtained in the nth diffracted order if the wavelength-shift ratio is equal to 1/n.
The Burt pyramid is an efficient algorithm that produces a multiresolution bandpass representation of image information. Since the pyramid is a potentially powerful tool for advanced television image processing, a pyramid processor has been designed and constructed to perform both the decomposition and reconstruction processes of the algorithm on digitized television signals in real time. The processor features a novel cascaded-filter pipeline architecture that achieves the performance of high-order filters through the repetitive use of low-order filters by resampling. In addition, each successive lower-spatial-frequency band is calculated from previously computed bands, while the linear sampling density is reduced by two per dimension to eliminate redundant information that is wasteful of both memory storage and processing time. Because the architecture is modular, highly structured, and arithmetically well-behaved, it is attractive for VLSI implementation.
An analysis is given of the effect of additive and multiplicative noise in the data and servo channels of both write-once and magneto-optic recording systems. Electronic (additive) noise is most severe in the magneto-optic data channel, and laser noise is of particular concern in both the sum tracking channel and the write-once data channel. In a typical optical memory system, the maximum laser light fluctuation should never exceed 10% of the average light level. In the differential channel of such a system, the laser noise should be further reduced by another order of magnitude. In the data channel, nonreturn-to-zero (NRZ) encoding has the greatest immunity to noise. Analysis is made of the extent to which the S/N margin is lost when various run-length-limited codes are used, for both additive and multiplicative noise.
An intercomparison has been conducted among three independent scales of spectral irradiance: two source-based and one detector-based. Specifically, a radiometer composed of a silicon photodiode, an interference filter, and an integrating sphere was characterized and calibrated against an absolute silicon detector standard at 600 nm using a cw dye laser. This radiometer was then used to measure the spectral irradiance at 600 nm from spectral irradiance lamps calibrated against a gold-point blackbody, and the spectral irradiance at the same wavelength from the NBS electron storage ring, SURF-II. Intercomparisons of this type are an important check of the agreement between these independent radiometric techniques. It was found that the detector scale indicated a spectral irradiance at 600 nm that was 0.76% lower than predicted by the gold-point blackbody scale and 0.25% higher than predicted by the electron storage ring scale. This result implies agreement within the overall quadrature uncertainties of ±0.25% for the detector scale, ±0.84% for the gold-point blackbody scale, and ±0.60% for the electron storage ring scale.