Johnson and Branstetter favor Laguerre-Gauss quadrature over the usual series summation2 for the fast, precise evaluation of integrals of the Planck function. But the only disadvantage of the usual series, which is an expansion in powers of exp (-x), is that it does not converge very rapidly for small x. This problem is easily solved through the use, for small x, of a more rapidly converging expansion in powers of x. The two series together provide more precision than the Laguerre-Gauss quadrature for a comparable number of terms, and are fast and simple to use. This note summarizes the expressions needed to obtain any required accuracy with the two-series approach. Expressions are given which bound the fractional residual errors when the series are truncated; and to show the precision attainable, these bounds are plotted as functions of x for 5 and 10 terms of each series.
The "special issue" has been a characteristic of Optical Engineering for many years. Under Optical Engineering's last editor, Dr. John DeVelis, special issues came to be a highly effective way of drawing fields together and drawing the attention of the optics community to those fields. In this editorial, I want to affirm my desire to maintain and strengthen what John has started. The particular questions I want to address are:
Over one year ago, the concept of this special issue was first formulated. It was conceived, initially, as one special issue on digital image processing and a second special issue on optical data processing. These topics have been retained. However, rather than two separate issues, we have chosen to intersperse digital and optical papers throughout two issues. This has been done both for content and to emphasize the similarities in techniques and application areas between these two powerful computing disciplines. We trust that these truly joint-dual special issues will typify the rapidly increasing marriage of the disciplines of optical data processing and digital image processing.
Matched filtering is often too sensitive to within-class variations (rotation, translation, magnification, etc.) and too insensitive to between-class variations. Improvements can be sought by changing the input, changing the filter, or changing the way the data is handled. We review both systematic and ad hoc methods to accomplish those improvements. The resulting improvements make coherent optical pattern recognition far more versatile in 1980 than it was when it began in 1964.
The principle of synthetic aperture radar (SAR) image formation is reviewed in preparation for a discussion of both optical and digital processing techniques. The tilted-plane optical processing approach is presented as being representative of optical techniques. Since the newer digital approaches can take several forms, three classes of digital processors are examined: direct convolution, frequency multiplexing, and frequency analysis of dechirped data. A subjective listing of the relative merits for both processing media is presented. Both are found to be technically viable. The final choice will depend primarily upon the application requirements.
Computer image processing at Caltech's Jet Propulsion Laboratory (JPL) began over a decade ago with digital processing of single images of the lunar surface returned by the Ranger and Surveyor spacecraft. During the past decade, advanced techniques have been developed at JPL for processing large volumes of imagery returned by the more recent planetary spacecraft, including the Mariner 6, 7, and 9 missions to Mars, the Mariner 10 mission to Venus and Mercury, the Viking Orbiter and Lander Mars missions, and the current Voyager mission to Jupiter and Saturn. In addition, the Image Processing Laboratory (I PL) has become involved in processing of earth resources imagery acquired by the Landsat earth orbital satellite and a variety of other sensors flown on aircraft and spacecraft. The trend within the facility has been toward development of technology capable of processing increasingly larger image data bases. A variety of applications in both the planetary and earth observations areas involve merging and/or processing of more than one image and often require correlation of data acquired by a variety of sensors. This paper describes the evolution that has occurred, and the use of multiple imagery is illustrated through a variety of applications.
A hybrid optical/digital system is described for analyzing images based on textural information. A partially coherent white light optical processor is used to obtain the texture related information which is then processed digitally. Specifically the optical system generates a pseudocolor encoded image where the image color is a function of the local spatial frequency content of the image. Cluster analysis is the used to identify different color/texture regions. Experimental results are shown where the system has been applied to the problem of identifying distinct cloud types from satellite photographs.
A method for characterizing scene content from aerial images is presented. The method is demonstrated for a building complex scene for which a three-dimensional data base and corresponding aerial images were available. Intuitively, the complexity of the building scene as viewed in a projected image is proportional to the number of vertices visible in the view. The greater the number of vertices, the greater the complexity of the scene. To automate this approach, one must automatically locate vertices from aerial images of the scene and determine relations among the vertices. Objective measures of scene content should not only basically agree with the intuitive measures, but also possess certain desirable mathematical properties. Two such measures, structural entropy and structural content, which were previously developed, are applied to the building scene, and experimental results which illustrate the variation of these measures with range, azimuth, and elevation are provided. One application of the scene content measures is the prediction of overall scene content characteristics performance support for map matching systems. To illustrate this application, an error analysis is presented of the mean square error in the transformational computation between a three-dimensional scene and the corresponding two-dimensional projected images, given a number of corresponding vertices. The analysis illustrates that the best possible performance depends heavily upon the vertex location accuracy.
A new general multi-dimensional formulation of optical signal processing systems is advanced. All existing optical signal processing systems are described and summarized in this new notation. The advantages and disadvantages of the various systems are apparent from this formulation as well as a myriad of new system architectures and necessary component research.
Because of the growing number of digital data sources-from satellites such as the LANDSAT series to the increasing use of all-digital information distribution networks in the publishing industries-there has been a corresponding explosion in the need to store that data permanently at very high densities such that it can be retrieved conveniently and quickly. In parallel with these developments, the capability of high density optical data storage and retrieval techniques such as holography and optical spot recording to meet these needs has improved as well. Offering data transfer rates of up to 100 Mb/sec (spot recording) to over 1 Gb/sec (holographic) and information storage densities in excess of 100 Mb/in2, these technologies have now made it possible to configure complete systems for a wide range of mass storage requirements. The basic technologies and building blocks required for such systems, the general criteria for converting a mass memory specification to an optical storage system based on it, and examples of both holographic and direct spot systems are described. The present status of the Wideband Holographic Recorder System, with data transfer rates of over 1 Gb/sec, and the MASTAR 1015 bit archival mass memory with information storage densities of over 100 Mb/in2, are also reviewed.
The Air Force is actively supporting scientific research in optical processing through the award of contracts and grants to university and industrial research laboratories. The overall objective of this Air Force research program is to increase the flex-ibility of optical processors to the point of being able to perform any operation that is suitable for parallel processing. This paper will describe the currently active efforts in this program.
The design, fabrication and performance of two Wolter-Schwarzschild grazing incidence optics are described. Both telescopes have been figured by single point diamond turning and have achieved better than 15 arcsecond on-axis imaging. The telescope for the stellar spectrometer is an f/10 Type II system with an effective area of 225 cm2 at 250 A and 300 cm2 at 500 A. the primary has a maximum diameter of 38 cm and was fabricated in three elements. The copper-plated aluminum substrate was diamond turned; following nickel plating, the surface was polished and coated with evaporated gold. The performance during a sounding rocket flight is discussed. The pro-totype telescope for the Extreme Ultraviolet Explorer is an f/1.24 Type I system with an effective field of view of 5.0° diameter. The telescope has a maximum diameter of 40 cm and was fabricated as a single element. The aluminum substrate is to be diamond turned; the nickel plated surface will be polished and electroplated with gold. The design choice and defocusing optimization aimed at maximizing the field of view and number of image pixels is examined.
Presented herein is a class of satellite deployment options that exploit ATH (above-the-horizon) viewing by multiple satellites to achieve a significant impact in dim target detection. The deployment options can allow reduction of focal plane complexity and improve the global coverage properties of systems of satellites employing two-color (or two-bandwidth) detectors on a single focal plane. It is shown that six to ten satellites at 10,000 to 20,000 nmi altitude will provide global ATH coverage for target altitudes between 40 and 200 nmi, and that considerable multiple satellite coverage of targets will occur for targets at higher latitude. Exemplary system design applications are discussed.
The goals, design and performance characteristics of the Solar Flux Radiometer flown on the Large Probe of the Pioneer-Venus Multiprobe spacecraft launched on 8 August 1978 are described. Radiance measurements of the Venusian atmosphere in several spectral channels between 400 and 1800 nm as a function of altitude were made to further understand the role of solar radiation in the thermal balance of the atmosphere. Elevation and azimuthal measurements on the radiation field were made with five optical channels. Twelve filtered Si and Ge photovoltaic detectors were maintained near 30 C with a phase-change material. The detector output currents were processed with logarithmic transimpedance converters before being multiplexed and digitized with an 11-bit A/D converter. Atmospheric sampling in both elevation and azimuth was done according to a Gaussian integration scheme. The data output was serial digital at an average rate of 20 bits/sec and included housekeeping (sync, spin period, sample timing and mode). The received data were used to determine the deposition of solar energy in the atmosphere of Venus between 67 km and the surface along with upward and downward fluxes and radiances with an altitude resolution of several hundred meters.
A practical high-speed image information preprocessor using a channel plate image intensifier, which can process optical images, has been developed. The preprocessor can perform logical operations such as summation, subtraction, or multiplication of two neighboring frame images on 16 mm film and can also detect moving objects by processing sequential frames. The operation time can be as short as 20 nsec. The image brightness can be as low as about 10-4 1m/m2.
Aerial reconnaissance optical systems are often subjected to extremes in thermal environment. Designers have been able to engineer systems which maintain stable performance over a wide range of steady-state temperatures. But how long does it take a reconnaissance system to reach steady state after being subjected to a thermal shock (or step function)? This paper reports on experimental measurements made on several long focal length reconnaissance lenses exposed to thermal step functions. The most significant finding showed that the thermal recovery time is on the order of hours .... a period during which the reconnaissance mission may already have been completed.
In many production chains visual inspection of products is an important manufacturing consideration with respect to quality control. Recent progress in image processing and pattern recognition led the way to economically justified applications; modern technology enables the construction of such auto
mata featuring high reliability and constancy. This paper describes an automatic visual inspection machine performing intelligent control tasks in a very short time. The possible applications include dimension control of products; inspection of objects on shape, greyness, or texture; sorting of objects; positioning; etc. The design is based on a fast synthesis procedure of a TV image by profile extraction. The features to be controlled are extracted from these profiles and compared with upper/lower limits obtained from a learning process. The system performances are expressed by the elaboration of an industrial case, the real-time visual inspection of reed switches. More than 35 features are detected and controlled for each switch within 1 second. Several experiments with a prototype version have proven the feasible use as an industrial on-line controller.
The absorption spectra of air pollutant gases have a complicated structure due to vibrational and rotational levels of the polyatomic molecules. The Fourier transform of the absorption spectra from a gas mixture shows the characteristic peaks for each component gas. The error due to wavelength jitter of the scanning laser is effectively reduced by a Fourier-transformed least-squares method. This process is applied to air pollution measurement.
Very high spatial frequency irregularities with small amplitude on a polished surface that usually cause wide-angle scattering in conventional optical systems produce narrow-angle scattering in x-ray systems. Yet, even this small amount of scattering degrades the imaging properties of x-ray systems. Statistical surface perturbations and their effects on x-ray image quality are discussed. A model of surface roughness that is reasonably consistent with surface profile measurements and visible scatter profile measurement is discussed. This model is then used to evaluate the image profile of an x-ray system.
The design, construction, and performance of state-of-the-art germanium and germanium-diamond bolometers operated at 4.2, 2.0, 1.2, and 0.3 K will be discussed. These detectors have a broad range of applications, and are particularly important for the long wavelength (far infrared to millimeter) regions. Current results may be extrapolated to lower temperatures and higher levels of performance. A system which will operate at 0.1 K using a He4-He3 dilution refrigerator is under development.