Search

This paper is concerned with an application of a multilayer feedforward neural network for the vision detection of industrial pictures, and introduces a high characteristics image processing and recognizing system which can be used for real-time testing blemishes, streaks and cracks, etc. on the inner walls of high-accuracy pipes. To take full advantage of the functions of the artificial neural network, such as the information distributed memory, large scale self-adapting parallel processing, high fault-tolerance ability, this system uses a multilayer perceptron as a regular detector to extract features of the images to be inspected and classify them.

The pressure to "push the polymer envelope" is clear, given the exploding range of demanding applications with optical components. There are two keys to success: 1. Expanded range of polymers with suitable optical properties. 2. Sophisticated manufacturing process options with an overall system perspective: -Tolerances and costs established relative to need (proof-of-concept, prototype, low to high volume production). -Designed to integrate into an assembly that meets all environmental constraints, not just size and weight, which are natural polymer advantages. (Withstanding extreme temperatures and chemical exposure is often critical, as are easy clean-up and general resistance to surface damage.) -Highly repeatable. The thesis of this paper is that systematically innovating processes we already understand on materials we already know can deliver big returns. To illustrate, we introduce HRDT1, High Refraction Diamond Turning, a patent-pending processing option to significantly reduce total costs for high index, high thermal applications.

In this paper the performance of Hartmann-Shack wave-front sensor in the aberration measurements of human eyes is investigated and the accuracy and repeatability of measurements are presented qualitatively and quantitatively.

Predictive Lossy Compression has been found to be an interesting alternative to conventional transform coding techniques in multispectral image compression. Recently, High Efficiency Video Coding (HEVC) standard has shown significant improvement over state of the art transformation based still-image coding standard. In this paper we study the properties of multispectral image and propose a predictive lossy compression scheme based on HEVC. Empirical analysis shows that our proposed method is superior to the existing state of the art predictive lossy compression schemes.

Reproducing a natural and real scene as we see in the real world everyday is becoming more and more popular. Stereoscopic and multi-view techniques are used for this end. However due to the fact that more information are displayed requires supporting technologies such as digital compression to ensure the storage and transmission of the sequences. In this paper, a new scheme for stereo image coding is proposed. The original left and right images are jointly coded. The main idea is to optimally exploit the existing correlation between the two images. This is done by the design of an efficient transform that reduces the existing redundancy in the stereo image pair. This approach was inspired by Lifting Scheme (LS). The novelty in our work is that the prediction step is been replaced by an hybrid step that consists in disparity compensation followed by luminance correction and an optimized prediction step. The proposed scheme can be used for lossless and for lossy coding. Experimental results show improvement in terms of performance and complexity compared to recently proposed methods.

Improving the understanding of the optical scene components associated with coral reef imagery will advance the ability to map and monitor coral reefs using remote sensing. One tool that can aid in understanding the components in these scenes is the NIST Hyperspectral Image Projector (HIP). In this paper a hyperspectral scene is reformatted for projection using the HIP by first unmixing image spectra into endmembers. The abundance images representing each of the endmembers are then projected using the NIST HIP and collected by a hyperspectral imager. Since the scene is from a digital source, it can be used repeatedly without concern for changing measurement conditions. This work represents one of the first steps in developing scene projection capabilities that can be used for sensor characterization, algorithm testing or to have optical components changed independently in order to better understand the overall effects on the total observed scene.

Probe guides are surgical fixtures that are rigidly attached to bone anchors in order to place a probe at a target with high accuracy (RMS error < 1 mm). Applications include needle biopsy, the placement of electrodes for deep-brain stimulation (DBS), spine surgery, and cochlear implant surgery. Targeting is based on pre-operative images, but targeting errors can arise from three sources: (1) anchor localization error, (2) guide fabrication error, and (3) external forces and torques. A well-established theory exists for the statistical prediction of target registration error (TRE) when targeting is accomplished by means of tracked probes, but no such TRE theory is available for fixtured probe guides. This paper provides that theory and shows that all three error sources can be accommodated in a remarkably simple extension of existing theory. Both the guide and the bone with attached anchors are modeled as objects with rigid sections and elastic sections, the latter of which are described by stiffness matrices. By relating minimization of elastic energy for guide attachment to minimization of fiducial registration error for point registration, it is shown that the expression for targeting error for the guide is identical to that for weighted rigid point registration if the weighting matrices are properly derived from stiffness matrices and the covariance matrices for fiducial localization are augmented with offsets in the anchor positions. An example of the application of the theory is provided for ear surgery.

On board measurement of attitude position, for example roll angle, of autonomous vehicles is critical to the execution of a successful mission. This paper describes a real-time technique, which combines a polarization scanning reference source and a priori knowledge of the scanning pattern. Measurements in an anechoic chamber, as well as, field tests in a busy parking lot, verify the efficacy of the technique, for both line of sight and non-line of sight capability.

Several array processing architectures have been devised for mitigation of multipath interference. Those algorithms have focused on rejection of terrain scattered interference, sidelobe and mainlobe clutter, and hot clutter. Two adaptive array architectures for rejection of broadband coherent interference are developed in this paper. Both techniques involve uniform subbanding using the Discrete Fourier Transform or non-uniform subbanding using the wavelet transform. In the first approach, adaptation and weight computation is performed independently in each subband, and in the second technique the Frost constrained LMS algorithm is applied to all subbands in the transform domain. Due to the correlated nature of the jamming signal, spatial averaging is utilized in both cases. The mitigation performance of both algorithms is compared for various scenarios of coherent broadband interference. This paper also focuses on evaluating and rejecting multipath interference due to propellers or rotor of the radar receiver aircraft or helicopter. This type of interference is characterized by an induced doppler spread proportional to the angular velocity of the propeller or the rotor. Furthermore the reflected signal represents the near-field component of the propeller scattering signature. Spatial averaging followed by a transform-domain adaptive beamformer based on the Frost algorithm is used to reject this type of propeller generated multipath interference.

Current DICOM establishes connectivity between devices of multiple vendors defined by the storage and query/retrieve service classes. In today's IMAC systems imaging modalities are only sub-optimal integrated with other systems in the network. There is a need for higher integration of modalities with information systems like RIS and HIS. There is also a need for a higher integration of modalities with IMAC (PACS) systems. These requirements can be gained by the evaluation of dozens of Siemens PACS installations worldwide. It is not sufficient only to implement highly advanced functionalities into the PACS components, also imaging modalities and information systems, where most of the operator interactions takes place, must be integrated into an overall system based on common standards. CEN (Comite Europeen de Normalisation) TC 251/WG4 in cooperation with ACR-NEMA WG VI installed two project teams of experts working on the items mentioned above. The results were 2 DICOM/MEDICOM supplements: Supplement 8: `Storage Commitment Service Class' and Supplement 10: `Basic Worklist Management Service Class (Modality Worklist SOP Class)'. Additionally, technical reports were created, e.g. `VIEWS', a paper outlining future concepts for image management. A step towards higher integration and interoperability is achieved with the 2 DICOM supplements. Modalities can efficiently retrieve patient and procedure information from the information system. The operator on the modality can select scheduled procedure steps. The images sent to IMAC systems contain valid identification data. The additional protocol defined by the storage commitment ensures and double-checks, that the images of each procedure are reliably transferred to the IMAC system.

Temporal phase-shifting speckle interferometry (TPSSI) is a technique based on the recording of a sequence of speckle interferograms throughout the entire deformation history of an object. To calculate the phase map, a controlled phase change is introduced in the reference beam as a function of time. As temporal phase unwrapping involves only one-dimensional signals, this procedure is generally easier to carry out than 2-D spatial unwrapping. TPSSI also allows large-object absolute displacement fields to be obtained. This paper provides an overview of recent developments and applications of temporal phase-shifting speckle interferometry. These include the performance of different phase-shifting algorithms when a TPSSI system is used in the presence of harmonic and random vibrations and also for measuring mechanical vibrations. A novel adaptive phase-shifting algorithm is also presented that searches for the peak of the windowed Fourier transform of the modulated intensity signal and evaluates the phase at that frequency instead of at the carrier frequency. The presentation is illustrated with an application of TPSSI to the detection of sub-surface delamination defects in carbon fiber specimens submitted to vacuum loading. It is shown that re-referencing the temporal phase unwrapping algorithm after different time intervals allows the measurement of time-varying displacements in excess of 100 μm without the limit normally imposed by speckle decorrelation.© (2003) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

As visualization is applied to larger data sets residing in more diverse hardware environments, visualization frameworks need to adapt. Rendering techniques are currently a major limiter since they tend to be built around central processing with all of the geometric data present. This is not a fundamental requirement of information visualization. This paper presents Abstract Rendering (AR), a technique for eliminating the centralization requirement while preserving some forms of interactivity. AR is based on the observation that pixels are fundamentally bins, and that rendering is essentially a binning process on a lattice of bins. By providing a more flexible binning process, the majority of rendering can be done with the geometric information stored out-of-core. Only the bin representations need to reside in memory. This approach enables: (1) rendering on large datasets without requiring large amounts of working memory, (2) novel and useful control over image composition, (3) a direct means of distributing the rendering task across processes, and (4) high-performance interaction techniques on large datasets. This paper introduces AR in a theoretical context, provides an overview of an implementation, and discusses how it has been applied to large-scale data visualization problems.

In this paper, we implemented a holographic optical memory system which can store and reconstruct many image types using a new input and angular multiplexing method. As a new input method, the phase information of an image is inputed in the recording material instead of the brightness information. To do so. we represented the images, which were captured with a CCD camera or displayed on a computer monitor, on a liquid crystal television(LCTV) without a polarizer/analyzer. Therefore, we could achieve a uniform beam power regardless of the total brightness of the input image. and could apply a conventional scheduled recording method for the determination of the recording time for all the images needing to be stored. The reconstructed image was acquired by transforming the phase information into brightness information using an analyzer. The reference beam was generated by a Fourier transform of a binary phase hologram(BPH) which was designed with a simulated annealing(SA) algorithm on a LCTV. The LCTV was then interfaced with a computer. The proposed optical memory system is stable because the incident angle of the reference beam is controlled electronically by computer. We demonstrated the system using an optical experiment which stored and reconstructed various types of images in a 45° cut BaTiO 3 using the proposed holographic memory system.

It is well established that training backpropagation networks with noisy training sets increases the generalization capabilities of the network. Since input set noise is somewhat analogous to faults in the network, networks trained on noisy inputs should exhibit fault tolerance superior to that of similar networks trained on non-noisy inputs. This paper presents results of a study to determine the effect of noisy training sets on fault tolerance. Backpropagation was used to train three sets of networks on 7 X 7 numeral patterns. One set was the control and used noiseless inputs and the other two used two different noisy cases. Several network examples were trained for each of the three cases (no noise, 10% noise, and 20% noise). The noise was injected into each training image uniformly at random, and took the form of toggled (0 to 1 and 1 to 0) pixel values in the binary input images. After learning was complete, the networks were tested for their fault tolerance to stuck-at-1 and stuck-at-0 element faults, as well as weight connection faults. The networks trained on noisy inputs had substantially better fault tolerance than the network trained on noiseless inputs.© (1992) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

To inimize confusion over terminology, this paper commences by defining little known, controversial, and the author's terms. In particular it defines subjective, metric, and signature readers, listing different types of each. Accuracy and speed are briefly discussed.

The presentation on a two dimensional surface of a single point most descriptive of the values of multiple .uncorrelated parameters of an attribute has presented formidible problems if more than three measures are available. With only three measures, conversion to percentile scores and plotting on triangular cross section paper is frequently adequate. With more than three, involved computational procedures such as multivariate analysis have previously been necessary.

Coherent optical systems are useful in performing, two major operations in the field of data reduction: spectral analysis and linear filtering. The latter operation includes convolution, auto-correlation and cross-correlation. This paper emphasizes some new developments which have bearing on data reduction, while referring only briefly to the theory behind coherent optical systems. Of particular interest is the development of a technique which allows more general operations by linear filtering., Examples of spectral analysis and character recognition will be presented.© (1964) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Conventional data analysis presentations of complex data often result in visualization and com­ prehension problems. A method is required to present data in a form which is quickly and precisely under­ stood by the analyst.This paper describes some of the present data synthesis requirements. Motion picture presentation will be suggested as a possible solution. The prob­ lems and techniques for the production and utilization of motion pictures for this purpose will be discussed.Gary Smith is currently in the Master Business Administration program at St. Louis University. He is employed by McDonnell Aircraft Corp. and is assigned to the Flight Test Development Studies and Special Projects Group. He has held positions responsible for instrumentation of F-4B and F-4C flight test aircraft at St. Louis and Edwards Air Force Base.

The three dimensional impression of the real world is obtained from a variety of clues such as eye accomodation (change of focal length of eye), shading, motion of object, parallactic displacement of the views of the eyes and relative size of known objects. The computer controlled stereo optical display generation system that I will describe in this paper utilizes parallactic displacement of the two views presented to the eyes to give a subjective three dimensional impression of a displayed object.

ELECTRONIC SYSTEMS COMMANDPHOTO -OPTICAL & ASSOCIATED PROGRAMSMajor General J. S. O'NeillCommander, Electronics Systems DivisionUnited [3tates Air ForceL. G. Hanscom FieldWe in the Electronic SystemsDivision are pleased to co- sponsorthis seminar with SPIE. Though thisis our first co- sponsorship, we havebeen well acquainted with the work ofyour society since it was foundedeleven years ago.For this seminar on "FilmedData and the Computer," you haveselected a topic with far morequestions than answers. The manytechnical papers to be presentedtoday and tomorrow should help toprovide some of the needed answers.And we at ESD are very much interestedin these answers because photo -opticsare playing an ever increasing rolein the development of Command andControl Systems.ESD's program most closelyrelated to the theme of this seminaris the Optical Surveillance Subsystem

A Computer program RAW has been developed which permits a user to dynamically control the process of digitizing certain kinds of filmed images. A brief general rationale for film reading is followed by descriptions of the digitizing equipment and the program. Some conclusions are deduced. The paper concludes with a few brief guesses about the future of such activities.© (1966) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

An equipment is described which will automatically locate and then read off data points from CRT traces recorded transversely on 35 mm film. Up to 512 data points may be taken from each trace and the output, presented in binary form on 8 level paper tape is a measurement of the vertical distance of each point from a known base line. It has a print out accuracy of 1 part in 127, and an average sampling rate of 85 points per second.© (1966) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Hyperspectral and polarimetric data contain spectral response information that provides detailed descriptions of an object. These new sensor data are useful in automatic target recognition applications. However, such high-dimensional data introduce problems due to the curse of dimensionality, the need to reduce the number of features used to accommodate realistic small training set sizes, and the need to employ discriminatory features and still achieve good generalization (comparable training and test set performance). In this paper, we evaluate both hyperspectral and polarimetric feature sets and identify features useful for distinguishing targets from background. Various feature selection algorithms are assessed in terms of the goodness of the selected features and computation time. Our results show that (1) the integration of branch and bound algorithm and floating forward selection algorithm is promising for hyperspectral and polarimetric target detection applications; and (2) the combination of both hyperspectral and polarimetric features yields significantly better classification results than either hyperspectral or polarimetric features alone.

Two years ago at the tenth SPIE Symposium I presented a paper (Ref. 1) which illustrated some of the work JPL has been doing in the development of hardware and techniques for processing of the Ranger and Mariner pictures. It is the purpose of this paper to bring you up to date on our more recent activities.© (1967) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Many image processing problems, including the restoration of atmospherically degraded images , can be benefited by the combined use of digital computers and coherent optical equipment, preferably in real time. Various questions then arise: What types of operations can better be done on a digital computer than by coherent optics, and vice versa? How can one transfer data quickly between a digital computer and coherent optical equipment? In this paper, we first briefly review and then compare digital computer and coherent optical image processing techniques to indicate their relative merits and faults. Then, we propose schemes for data transfer between digital computers and coherent optical equipment, and discuss the hardware available for carrying out these schemes. Finally, we look into the future and describe some desirable features of large image processing centers.© (1967) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

An analysis of the performance of a cathode-ray-tube flying-spot-scanner film reading system requires a treatment of the gray level resolution of the system. This paper discusses three types of gray level encoding schemes for a digital film reading system; equal transmission, equal density and equal reliability levels. The maximum number of detectable levels, or the minimum gray level change detectable, to a given reliability is presented along with a trade-off analysis and a discussion of the utility of each method. Calibration and measurement of a hardware system by analysis of maximum signal available is discussed.© (1967) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

The formation of any hologram involves the exposure of a light-sensitive material to the holographic pattern resulting from the interference of object and reference beams. This recording process and the subsequent reconstruction of the resulting hologram are the essential steps of holography. They were considered carefully by Gabor (Ref. 1) and have since been examined theoretically and experimentally by many authors. It is the purpose of this paper to present a broad survey of the effects of hologram recording media, and to summarize in compact form the potential and actually observed properties of the various types of holograms resulting from the use of the different recording media. We shall also summarize the effects upon the reconstructed image of the detailed characteristics of some particular recording media. We shall concentrate on results, with few derivations of mathematical formulae or descriptions of experimental procedure.

This paper will describe the instrumental details of a modulation transfer function analyzer capable of presenting automatically the MTF curves in the frequency range of : a) 0.4 cycles/mm to 10 cycles/ mm; b) 4 cycles/mm to 100 cycles/mm; and c) 24 cycles/mm to 600 cycles/mm. The instrument is designed to produce an MTF curve,after the optical element has been set in place, in approximately 5 seconds, with an accuracy of ±5 percent of the indicated response value. This instrument has been used extensively for testing of lens systems, fiber optics components, and such electro-optical systems as image converters and image intensifiers. Pertinent details of the instrument setup for these various applications as well as data on numerous devices will be presented. Instrument accuracies and error analysis will be discussed along with details of y-axis normalization.© (1968) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This paper presents the ground work carried out to achieve automatic ne grained recognition of stone masonry.This is a necessary rst step in the development of the analysis tool. The built heritage that will be assessedconsists of stone masonry constructions and many of the features analysed can be characterized according to thegeometry and arrangement of the stones. Much of the assessment is carried out through visual inspection. Thus,we apply image processing on digital images of the elements under inspection. The main contribution of the paperis the performance evaluation of the automatic categorization of masonry walls from a set of extracted straightline segments. The element chosen to perform this evaluation is the stone arrangement of masonry walls. Thevalidity of the proposed framework is assessed on real images of masonry walls using machine learning paradigms.These include classi ers as well as automatic feature selection.Keywords: Built heritage analysis, ne grained visual categorization, classi er, feature selection

GML is an XML encoding for the modeling, transport and storage of geographic information, which has been making a significant influential on the ability of organizations to share geographic information among their Web based GIS applications. Ontology has been acknowledged to be the kernel methodology for capturing and sharing semantics of spatial information. This paper proposes a framework for integrating spatial information based on GML schema matching by using ontology technologies, in which GML is adopted as the common format for spatial information wrapping and mediation, and ontology is used to overcome the heterogeneity when matching different GML schemas. A prototype is also implemented based on the proposed framework.

During the past decade the ability to acquire, process and utilize technical and scientific information has undergone an explosive growth. This expansion is primarily due to the development of the large scale electronic digital computer. It is ironic to note however that the principal input to these magnificent processing systems is still manual. Classical methods of computer entry have originated in punched paper tape, and tabulator cards, both primarily prepared by manual methods. Only recently. have direct electrical and optical inputs become practical. Equipments such as the Program. Controlled Film Reader/Recorder described here offer the opening of a vast new source of scientific data to automated processing.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This paper discusses the use of multiple gratings to obtain additional information from photographs of spectra whose zeroth order is resolved in two dimensions. Techniques of microdensi-tometry are discussed. Data processing and computer programming methods are presented. Several configurations of data display are feasible. Results using parallel, perpendicular, and circular spectra arrays are given. The various conditions in which the images can be disentangled are derived. Some consideration is given to various applications.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Image processing with in-line optical systems has been improved with improved experimental techniques and with precise methods of filter generation. Both analog and digital processing techniques are available and combined hybrid processing methods have been developed. Analog processing includes in-line and holographic optical processing and hybrid processing utilizes the combined capabilities of digital and analog systems. This paper is concerned with in-line optical processing and one hybrid scheme utilizing digital processing for filter fabrication and an in-line analog system for image processing. Experimental results are demonstrated with illustrations including restored defocused and blurred imagery.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

ACKNOWLEDGMENTThe author wishes to thank ProfessorHerbert Freeman of New York University forhis patient guidance in the research thatled to this paper. This research wassponsored by the Air Force Office ofScientific Research, Directorate of Infor-mation Sciences, under Grant AF- AFOSR -68-1367, and is in partial fulfillment of therequirements for the degree of Doctor ofPhilosophy in the School of Engineering andScience of New York University. REr EHENCES 1. Roberts, L.G., "Machine Perception ofThree- Dimensional Solids," M.I.T.Lincoln Laboratories Technical ReportNo. R 315, May 1963. 2. Smith, A. F. , "A Method for ComputerVisualization," M.I.T. ElectronicSystem Laboratory Technical Memorandum, No. 8+36 -TM -2, September 1960. 3. Rabinowitz, Andrew D., "An Approach tothe Problem of Reconstructing Polyhedrafrom Two or More of Their PerspectiveProjections,'" N.Y.U. Technical ReportNo. 400 -181, April 1968.The use of computers to analyze perspective projections, such asphotographs, and obtain three -dimensional geometric descriptions of thesubjects, has many obvious applications. Obtaining such data from two ormore perspective projections eliminates many of the ambiguities and omissionsof data inherent in obtaining such data from a single projection. howevera new problem, that of correlating the images in the different projectionsis introduced. In order to cope with this problem, a perspective projectionsubsystem is defined for each center of projection and the correspondingpicture plane. Then these modules are considered in several possibleconfigurations. Different techniques based on geometric properties andand physical constraints are used to determine the necessary correlationsfor the various configurations. In addition, different types of perspectiveprojections are considered separately. Experimental results from programswritten for a CDC -6600 digital computer will be described.

The usual practice in pattern recognition is first to adopt some working definition for the patterns of interest and then to expend most of the effort on the equipment and/or automated techniques required to find and recognize these patterns. This paper is concerned only with the first step; it outlines a strategy, based on psychophysical methodology, for determining those features of the pattern which are most likely to be informative. The approach used is termed psychopictorics; this is defined as a subfield of psychophysics which is concerned with pictorial stimuli, and in which it is assumed that, in a picture, the information of significance to the human observer may be characterized and analyzed in terms of the properties of perceived objects. The analysis of these properties involves the measurement of many psychophysical variables while the observer is responding to repeated, controlled changes of the features of single objects in the picture. Thus psychopictorics is strongly dependent on the development of computer picture processing techniques which permit such controlled manipulations without unduly degrading the quality of the picture.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This paper describes diffraction-pattern sampling as a basis for automatic pattern recognition in photographic imagery; it covers: diffraction-pattern generation, diffraction-pattern/image-area relationships, diffraction-pattern sampling, algorithm development, facility description, and experimental results which have been obtained over the last few years at General Motors' AC Electronics-Defense Research Laboratories in Santa Barbara, California. Sampling the diffraction pattern results in a sample - signature - a different one for each sampling geometry. The kinds of information obtainable from sample-signatures are described, and considerations for developing algorithms based on such information are discussed. A tutorial section is included for the purpose of giving the reader an intuitive feeling for the kinds of information contained in a diffraction pattern and how it relates to the original photographic imagery.© (1969) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

A statistical analysis of degradation in a scanned image has been carried out by Beall. (Ref. 1) In the above analysis, the mean-squared value of the error function between the object and image was used as a measure of the degree of fidelity in the image, and numerical data have been published for several particular cases. A similar approach to image transfer in optical fiber bundles is presented. In this paper, the one-dimensional line-scan analysis proposed by Beall is extended to encompass a two-dimensional sampling process for evaluating the image transfer in a fiber bundle. Interesting results occur when the evaluation technique is extended to the comparison of the imaging properties of a static bundle and a bundle which is scanned dynamically. (Ref. 2) This analysis indicated that there is a significant difference in resolution between the static case and the dynamically-scanned case. The difference is found to be largely determined by the fiber configuration within the bundle. Experimental results in the case of the static bundle will be presented.© (1970) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Optical sensors are rinding increased usage in inspection and quality assurance applications. The versatility and high speed data gathering capabilities of imaging sensors lend themselves to complex applications such as pattern recognition and gray-scale defect deter-minations. Applications of this type usually require computer capability to mandLutte and digest the data but optical scanning rates usually outpace the instruction cycle rates of microcomputers. This paper discusses a method of handling high speed imaging data to pro vide real time computer analysis, decision and control. Methods and technical designs are described that provide data compression, formatting and preprocessing. Hanipulative tech.- niques for both digital and analog (video) signals are discussed.

Reseachers seeking a histological determination of microscopic specimen morphology face the basic problem of analyzing three-dimensional microscopic specimens with two-dimensional imaging devices. This tedious task can be done at the microscope but reproduction of images in hard copy or digital form entails the loss of depth information. Quantitative measurements must generally be made on 2-D images. Several methods have been proposed and implemented to obtain and display 3-D images. In this paper we discuss an approach combining digital image processing techniques with stereo display techniques. This allows control of image quality and selection of different 3-D characteristics for display.

This paper is concerned with the automatic calculation of left ventricular volumes from a cineangiogram by a digital computer. It will first describe three digital techniques (1) to restore the original dimension of the objects by a logarithmic transformation, (2) to subtract images before and after the injection of radio opaque agent, and (3) to detect boundaries by a threshold method based on statistical principles and heuristics. Experimental results on several frames will be presented to demonstrate the feasibility of these techniques. Once the boundary of the ventricle is determined, the computation of the volume is performed by the usual Simpson's method.© (1971) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Automated 3D modeling of building interiors is useful in applications such as virtual reality and environment mapping. We have developed a human operated backpack data acquisition system equipped with a variety of sensors such as cameras, laser scanners, and orientation measurement sensors to generate 3D models of building interiors, including uneven surfaces and stairwells. An important intermediate step in any 3D modeling system, including ours, is accurate 6 degrees of freedom localization over time. In this paper, we propose two approaches to improve localization accuracy over our previously proposed methods. First, we develop an adaptive localization algorithm which takes advantage of the environment's floor planarity whenever possible. Secondly, we show that by including all the loop closures resulting from two cameras facing away from each other, it is possible to reduce localization error in scenarios where parts of the acquisition path is retraced. We experimentally characterize the performance gains due to both schemes.

The paper presents the design of a real-time and low-cost embedded system for image acquisition and processing in Advanced Driver Assisted Systems (ADAS). The system adopts a multi-camera architecture to provide a panoramic view of the objects surrounding the vehicle. Fish-eye lenses are used to achieve a large Field of View (FOV). Since they introduce radial distortion of the images projected on the sensors, a real-time algorithm for their correction is also implemented in a pre-processor. An FPGA-based hardware implementation, re-using IP macrocells for several ADAS algorithms, allows for real-time processing of input streams from VGA automotive CMOS cameras.

Image segmentation, a key component in many Automatic Target Recognition (ATR) systems, has received considerable attention in the research community in recent years. A variety of segmentation approaches exist, and attempts have been made to combine various approaches in order to find more robust solutions. In this paper, the authors describe an inference fusion architecture for combining individual segmentation concepts which results in improved performance over the individual algorithms. We consider segmentation algorithms with several disparate cost functions as experts with a narrowly defined set of goals. The information obtained from each expert is combined and weighted with available evidence using an agent based inference system, resulting in an adaptive, robust and highly flexible image segmentation. Results obtained by applying this approach will be presented.

In the computer processing of dynamic images (cinefilm, TV) data volume and rates are usually excessively high. This is because of the addition of two spatial demensions to the normal time-domain processing problem. It has therefore been necessary to reduce the spatial and intensity resolution of the converted images. This paper examines the errors resulting from these quantization trade-offs, as applied to several medical image processing problems.

The purpose of this paper is to describe the image processing research at MIT and to present some qualitative considerations about picture quality.© (1973) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

The paper describes a strategy we use to translate an existing conventional archive into a digital form. The method is directed to large archives comprising documents with essential graphic constituent (handwritten texts, photographs, drawings, etc.) that result in images. Our technology of image digitization, storage and presentation in multiple resolutions is specifically discussed. Main structural components of the digital archive are relational database and image bank, physically separated but logically linked together. The components make up three-level distributed structure consisting of primary archive, its regional replicas, and various secondary archives (among them subsets presented in the Web and CD/DVD-ROM collections). Only authorized user is allowed to access two upper levels, and the bottom level is open for free public access. A secondary archive is created and updated automatically without special development. Images in the bank are stored in multiple resolutions, and linking the proper image to a database record comes to be dynamical, dependent of user interaction context (e.g. channel bandwidth, user permissions, etc.) Such construction allows us to combine reliable storage, easy access and avoid intellectual property protection issues. We also presents several digital archives already implemented on this basis in the Archive of the Russian Academy of Sciences.

Optical spectral analysis is finding wider application in data processing. The major advantages of the technique are very high speed and high resolution (Refs. 1,2,3). As an example, the laser beam system considered in this paper can continuously record a signal of several MHz bandwidth with no loss of data. The laser beam recorder may be considered as the input channel of a computer. The data are stored on film in a compact raster format in analog form and can place about a half million samples in the aperture of the optical computer. Because the optical analyzer or computer produces output based upon this large data sample, the resolution of the system is greatly enhanced.

Optical data processing is one of the most powerful and versatile processing tools to emerge in recent years. Numerous texts, conferences, and special issues of journals have been devoted to its many accomplishments. There are many types of optical processors: some use non-coherent rather than a coherent light source, while others use non-holographic rather than holographic processing methods. The hybrid optical processor described in this paper can be partitioned into two sections. One is a conventional coherent optical processor while the second section is a digital minicomputer. These two sections communicate via a special purpose digital interface. With this configuration, the best features of both optical and digital processing are retained.© (1974) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

This paper provides a summary of various known techniques for generating holograms by means of a digital computer and associated display devices, and discusses a recently invented method which is some cases offers distinct advantages over alternate approaches.© (1974) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

One method of generating synthetic or computer generated holograms is to first compute the Discrete Fourier Transform (DFT) of the desired image. The hologram is then formed by coding in some manner on photographic film the complex values of this DFT. This paper will describe a method for generating synthetic binary holograms that avoids completely the necessity for computing the DFT of the desired image. The relationship between this new method and previously used techniques will be explored.© (1974) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.