This study deals with the problem of estimating a transmitted string X//s, when the only observable string is its noisy version Y. Y is assumed to contain substitution, deletion and insertion errors. Based on three ge...
详细信息
This study deals with the problem of estimating a transmitted string X//s, when the only observable string is its noisy version Y. Y is assumed to contain substitution, deletion and insertion errors. Based on three generalized edit distances of substitution, insertion and deletion and on the properties of the noisy channel through which X//s is transmitted, a distance D(X/Y) is defined between a word X contained in the dictionary H and the noisy string Y. The best estimate X* of X//s is defined as that element of H which minimizes this distance. Using dynamic programming principles, an algorithm is then presented which yields x* without computing individually the distances between every word of H and Y.
The objective of computer vision system is to outline the objects in a picture and label them with an appropriate interpretation. This study proposes a new paradigm for a modular computer vision system which is both d...
详细信息
The objective of computer vision system is to outline the objects in a picture and label them with an appropriate interpretation. This study proposes a new paradigm for a modular computer vision system which is both data directed and knowledge based. The system consists of three different types of units, two of which are associative data memories implemented as relational databases. The short term memory (STM) contains the raw color picture data and the most current interpretations and deductions about the original scene. The long term memory (LTM) contains a detailed model of the scene under consideration. A collection of analysis processors, each of which is specialized for a particular task can communicate with both of these memories.
Object location in computed tomography images is a preliminary step required for many automated measurements which may be useful in many diagnostic procedures. Most object location, imageprocessing techniques are eit...
详细信息
Object location in computed tomography images is a preliminary step required for many automated measurements which may be useful in many diagnostic procedures. Most object location, imageprocessing techniques are either globally based such as histogram segmentation or locally based such as edge detection. The method described uses both local and global information for object location. The technique has been applied to the location of suspected tumors in CT lung and brain images. Sorting and merging steps are required for eliminating noise regions but all suspected tumor regions have been located. Measurements such as boundary roughness or density statistics may also be made on the objects and used to further identify suspicious regions for further study by the radiologists.
A description is provided of some novel algorithms for patternrecognition research and a framework for efficient development, maintenance, and sharing of interactive software among several users and diverse applicati...
详细信息
A description is provided of some novel algorithms for patternrecognition research and a framework for efficient development, maintenance, and sharing of interactive software among several users and diverse application areas. This modular interactive software system (MISS) forms the basis of a general purpose image analysis and patternrecognition research system (IPS). The algorithms and some preliminary results are discussed first. Then, the MISS and IPS software systems are described.
Modern scanning techniques, such as computed tomography, have begun to produce true three-dimensional imagery of internal structures. The first stage in finding structure in these images, like that for standard two-di...
详细信息
Modern scanning techniques, such as computed tomography, have begun to produce true three-dimensional imagery of internal structures. The first stage in finding structure in these images, like that for standard two-dimensional images, is to evaluate a local edge operator over the image. If an edge segment in two-dimensions is modelled as an oriented unit line segment that separates unit squares (i. e. , pixels) of different intensities, then a three-dimensional edge segment is an oriented unit plane that separates unit volumes (i. e. , voxels) of different intensities. In this study the authors derive an operator that finds the best oriented plane at each point in the image. This operator, which is based directly on the 3-D problem, complements other approaches that are either interactive or heuristic extensions of 2-D techniques.
A system for detecting and classifying images on astronomical plates is described. In particular, the very faint and barely resolved images of 24th magnitude galaxies are shown to be detectable and distinguishable fro...
详细信息
A system for detecting and classifying images on astronomical plates is described. In particular, the very faint and barely resolved images of 24th magnitude galaxies are shown to be detectable and distinguishable from stellar images of similar brightness. The algorithms and other implementation considerations used in the construction of the Faint Object Classification and Analysis System, FOCAS, are given. Detection and classification at 24th magnitude is achieved with N//2 presensitized IIIaJ plates exposed for one hour in the KPNO 4-m telescope prime focus camera. For a digital 6000 multiplied by 6000 pixel image made from the central square area 37 ft on a side approximately 16,000 objects are found to 24. 7 mag on such plates. Since the resulting catalog of objects is 3 or 4 magnitudes deeper than existing surveys a variety of internal tests must be used to assess the accuracy and completeness of the list of objects. The tests and simulations made to support these claims are also discussed.
Future generation x-ray computed tomography scanners will be characterized by their ability to record simultaneously a sufficient number of x-ray projections to allow reconstructions of multiple adjacent cross section...
详细信息
Future generation x-ray computed tomography scanners will be characterized by their ability to record simultaneously a sufficient number of x-ray projections to allow reconstructions of multiple adjacent cross sections of the object under study. An ability to repeat the entire data collection procedure with great rapidity, allowing many scan passes per second, should encourage research and diagnostic studies of moving organs such as the heart and lungs in truly three dimensions and in real time. A combined series of algorithmic, software, special-purpose computer architecture, and hardware implementation studies have demonstrated significant progress toward computed tomography reconstruction processing rates of 10**9 to 10**1**0 arithmetic operations per second.
Presented are the details of a study conducted in the general area of patternrecognition. Broadly speaking, the objective was to identify ways by which a patternrecognition system could be implemented on a microcomp...
详细信息
Presented are the details of a study conducted in the general area of patternrecognition. Broadly speaking, the objective was to identify ways by which a patternrecognition system could be implemented on a microcomputer in a feasible way through a specific design. Such a design would strive to take advantage of mathematical equivalencies by which reductions in processing complexities would be achieved and provide a test bed for design tradeoffs and sensitivity studies. Automatic Speech recognition is broken into four sequential processing stages: data acquisition, preprocessing, feature extraction and classification. A microcomputer was selected to implement an all-digital version of an automatic speech recognition system based on the technique first developed by T. B. Martin. It is constrained to a restricted vocabulary and a limited set of users.
Objects are analyzed into generalized cylinders and relations between them. The class of cylinders used is restricted to straight-line axes (spines), elliptical cross-sections, and a sweeping rule whereby the lengths ...
详细信息
Objects are analyzed into generalized cylinders and relations between them. The class of cylinders used is restricted to straight-line axes (spines), elliptical cross-sections, and a sweeping rule whereby the lengths of the major and minor axes of the ellipses change linearly and independently. Most smooth elongated objects can be described as compositions of such cylinder primitives. Descriptions can be compared to each other and to object prototypes. Two algorithms are developed and tested for extracting cylinder descriptions from stacks of parallel slices. A local algorithm creates cylinders from elliptical regions on adjacent slices. A predictive algorithm hypothesises cylinders from regions on non-adjacent slices and then tests the hypotheses by intelligently requesting intermediate slices. Both algorithms explain complex non-elliptical regions (joints, typically) by using cylinders impinging upon them as keys to their shapes.
A recursive algorithm is given to directly compute the fast Fourier Transform (FFT) of a two-dimensional image array. The algorithm builds the FFT of a 2**m by 2**m matrix left bracket f right bracket by first mapping...
详细信息
A recursive algorithm is given to directly compute the fast Fourier Transform (FFT) of a two-dimensional image array. The algorithm builds the FFT of a 2**m by 2**m matrix left bracket f right bracket by first mapping its entries to form a new matrix left bracket f//1 right bracket , in a manner analogous to that used in one-dimensional FFT's. The FT of left bracket f right bracket is then found by computing the FT of all 2 by 2 submatrices of left bracket f//1 right bracket and recursively combining them into higher level submatrices until the 2**m by 2**m level is reached. This algorithm is shown to be computationally faster than the conventional method based on one-dimensional FFT's. The main features of this method, however, are its simplicity of updating the FFT of an image array when some of the array elements have been modified, and its suitability for implementation on a multiprocessor computer system. Applications of this algorithm to imageprocessing are discussed and illustrative examples are given.
暂无评论