Abstract: A system and method in accordance with the present invention includes a scanning assembly and a storage device coupled to a programmed computer with a set of instructions for carrying out an assessment of a document image. The system and method operate by: processing the document image to obtain one or more attributes related to the geometrical integrity of the document image; selecting a threshold value from a database for each of the obtained attributes; and then comparing each of the obtained attributes against the threshold value selected for the obtained attribute to determine a difference for each and then evaluating one or more of the differences using predetermined criteria to provide evaluation results of the geometrical integrity of the document image.

Abstract: A method for distinguishing photocopied or laser-printed documents from original documents produced by offset printing, handwriting, or typewriting. A document is scanned at low-resolution and at high-resolution to produce a low-resolution and a high-resolution matrix representation of the presence or absence of ink or toner at discrete locations on the surface of the document. Printed regions detected at low-resolution are used to mask regions of the high-resolution matrix representation from the analysis. The remaining unmasked regions of the high-resolution matrix representation are analyzed to detect discrete microdots uniformly distributed within those regions. The presence of microdots on the surface of the document indicates that the document was produced as a photocopied or a laser-printed duplicate.

Abstract: There are provided an image processing apparatus and method which can properly determine the number of quantization levels for quantizing image data by properly removing noise from the background of the image data. A histogram forming unit forms the output frequency distribution of densities of image data. A background density width detecting unit determines a density area exhibiting output frequency values equal to or larger than a predetermined threshold and including a maximum output frequency value in the formed distribution. A quantization level count determining unit decides the number of quantization levels on the basis of the number of maximum values included in the distribution other than the determined density area.

Abstract: Upon processing a visible light image signal and infrared image signal respectively obtained by illuminating a transparent document with light beams coming from a visible light lamp for mainly emitting visible light and an infrared lamp for mainly emitting infrared light, and photoelectrically converting optical images of the transparent document, a histogram is generated on the basis of the infrared image signal, a threshold value is calculated based on the histogram, and infrared image signal components equal to or smaller than the threshold value are extracted by comparing the calculated threshold value and infrared image signal components. Visible light image signal components corresponding to the extracted infrared image signal components are interpolated using surrounding visible light image signal components.

Abstract: Digital images often have undesired color casts due to unusual illuminant sources. A color correction method is provided that uses adaptive segmentation to identify the presence of such casts, estimate their chromatic strength, and alter the image's near-neutral-color regions to compensate for the cast. The method identifies most major objects in a scene and their average color. A color space having an achromatic axis is provided with a default gray radius that is used for establishing bins of neutral color in a chromatic plane. The bins are populated with the pixels of the digital color image based on pixel chromaticity. A color histogram computed with chromaticity peaks for the pixels in the bins within the default gray radius. The default gray radius is adjusted to a gray radius based on the chromaticity peaks in the color histogram.

Abstract: A method and apparatus for selecting colors for blocks for use in a Block truncation Coding scheme is disclosed. The method includes finding the largest cluster of pixels and selecting that color for the first color in the block. The average color of the remaining pixels is assigned to the second color. The invention may be applied in iterative fashion if more than two colors are to be assigned to the block.

Abstract: A money identifying method and a device which can identify money with high precision and quickly by optimally binarizing a surface image of the money. The center coordinates and radius of subject coin C are extracted by coin shape extraction section 106 from the surface image of the subject coin C obtained by area sensor 103, and an effective threshold calculation range is extracted by threshold calculation range extraction section 107 from the extracted center coordinates and radius of the coin. And, a density histogram is created by histogram generation section 108 on the basis of the threshold calculation range, an optimum binary threshold value is calculated from the density histogram by binary threshold calculation section 109 according to a discriminating analysis, and the optimum binary threshold value is used to binarize an image for identifying the subject coin C.

Abstract: Heretofore, an operator has been required to judge the type of image on a display and select any of various effect processings manually, thus automation has been impossible. By totaling luminance values of pixels into a luminance distribution which pixels have been selected by a thinning-out processing for image data (step S102) it becomes possible to count the number of colors used (step S104). As a result, if the number of colors used in image data is large, it can be judged that the image data is of a natural picture. Besides, a contrast enlarging process (step S110), a saturation highlighting process (step S112) and an edge highlighting process (S114), which are suitable for application to a natural picture on the basis of the results of such judgment, can be selected automatically while changing the respective highlighting degrees.

Abstract: A method and apparatus for providing media-independent security for a document may be programmed to create a document file having two or more components. In one embodiment, a document may include a background object, an image object (e.g. text, graphic, both, or the like), and a watermark object. When output, the image object is directly interpretable by a user. Meanwhile, in the background object, watermark object, or both, a high-resolution pattern may be stored to be output with all copies of the document. Encoded in some binary symbol in the pattern is security data. Resolution is high enough that the binary symbols are undetectable by a human eye. A processor may be programmed to recognize (e.g. read) the pattern, decode the pattern into binary data, and decode the binary data to characters directly interpretable by a user. Information relating to creation and control of a document, signature, or the like, may all be encoded independent from the principal image (e.g.

Abstract: The present invention relates to a multimedia query using a histogram, and more particularly, to a method for configuring a histogram which provides interoperability between histograms configured by non-uniform bin quantization using bits of a different number, and provides progressive bit processing for keeping a constant performance even when using the front parts of total bits of each bin value in the order of time series.

Abstract: A method, system and computer readable medium for automated detection of lung nodules in computed tomography (CT) image scans, including generating two-dimensional segmented lung images by segmenting a plurality of two-dimensional CT image sections derived from the CT image scans; generating three-dimensional segmented lung volume images by combining the two-dimensional segmented lung images; determining three-dimensional lung nodule candidates from the three-dimensional segmented lung volume images, including, identifying structures within the three-dimensional segmented lung volume images that meet a volume criterion; deriving features from the lung nodule candidates; and detecting lung nodules by analyzing the features to eliminate false-positive nodule candidates from the nodule candidates.

Abstract: A system and method suppresses automatically a background of a document being scanned. A histogram circuit generates a histogram from a predetermined group of scanlines of image data collected at a lead edge of the document being scanned, and a gain correction circuit calculates a gain value from the histogram and applies the gain to the image data. The histogram circuit generates a next histogram from a next predetermined group of scanlines of image data collected from the document being scanned. The gain correction circuit then calculates a next gain value from the next histogram. A comparator enables the application of the next gain value to the image data if the next gain value is less than a previously calculated gain value.

Abstract: An apparatus and method for automatically identifying faults in the operation of a machine vision measuring systems provides an improved self-diagnostic capability for machine vision based metrology and tracking systems. The method and apparatus validate performance of tracking operations, and detect deterioration that may be caused by electronic noise, environmental contamination, etc. A mathematical model of the target visualized by the system is created and stored. A target is imaged in the field and fiducials of the target are identified. Centroid positions of detected fiducials of the imaged target are compared to the centroid positions of fiducials in the mathematical model. When a fiducial is obscured or dirty, its geometric characteristics (such as centroid, brightness, edge smoothness, area, or shape) differ from the true or idealized values of the characteristics.

Abstract: A process and an apparatus for converting multi-valued image data obtained, for a plurality of pixels of a continuous tone image, by an image input device, into binary image data representing a halftone image to be output by an image output device. The conversion is performed by comparing the multi-valued image data for the plurality of pixels with threshold values, which are determined in advance so as to be specific to a combination of the image input device and the image output device. Alternatively, the multi-valued image data may be corrected before comparison with threshold values, by using correction data which is determined in advance so as to be specific to a combination of the image input device and the image output device.

Abstract: An image processing device such as an image scanner, etc., and its objective is to actualize a proper binarizing routine specific to the types of images. A histogram of density values is drawn based on decoded data, and after the respective levels and frequencies of the white and black peak values have been calculated, the types of decoded images are classified based on their relationships. After the original type has been thus determined, a binarization threshold value calculation method optimally suited for said type is selected, and after the threshold value has been thereby calculated, an image binarizing routine is executed.

Abstract: The present invention concerns an image processing device and an image processing method for processing a decoded image, and its objective is to optimally binarize a decoded original. After the respective density histograms of colors R, G, and B have been drawn, the hierarchical relationship between the low-density-side peak value and high-density-side peak value is investigated. A color component which is to become a binarization object is determined based on the comparison results on the number of color components whose high-density-side peak value is/are higher and the number of color components whose low-density-side peak value is/are higher, and the color of the pixel to be outputted is determined based on the comparison of the pixel with a threshold value.

Abstract: A device for and method of detecting motion between a reference image and a test image by acquiring the images; aligning the images; dividing the images into blocks; masking certain blocks; differencing corresponding blocks; median filtering the differences; low-pass filtering the outputs of the median filter; generating a normalized histogram for each output of the low-pass filter; generating a model of gaussian noise; calculating the distance between the noise model and each normalized histogram; comparing each distance calculated to a user-definable threshold; and determining if motion has occurred between the images if a certain number of distance calculations are at or above the user-definable threshold. If a scene is to be continuously monitored and no motion occurred between the previous reference image and the previous test image then a new test image is acquired and compared against the previous reference image as described above.

Abstract: Noise in digitized image data is reduced by providing an array of pixels for each of which a gray level has been determined. For each column of the array, a distribution of gray levels is derived, and a range of acceptable gray levels is set based on the distribution. For pixels with a gray level outside the range, the gray level is changed to reduce the influence of noise in the imaged data.

Abstract: A sample hold circuit is disposed which, in accordance with a hold signal, can selectively execute either of outputting a threshold signal output from an Automatic Threshold Control circuit, to a comparator as a threshold signal while allowing the voltage to be as it is, or outputting the signal while holding the voltage. Immediately before reception of synchronizing data in which signals of “1” and “0” are alternatingly repeated is ended, the voltage of the threshold signal is held by a sample hold circuit. Therefore, when effective data is received after the synchronizing data, the voltage of the threshold signal is constant. The comparator compares the voltage of the input signal with the voltage of the threshold signal used as a reference. Based on a result of this comparison, a digital signal is reproduced from the input signal.

Abstract: A method and apparatus are disclosed for constructing a virtual microscope slide comprised of digitally scanned images from a microscope specimen. The digitally scanned images are arranged in a tiled format convenient for viewing without a microscope, and for transferring the tiled images for viewing by another at a remote location. Several original microscope views at a low magnification are digitized and stored as digitized images coherently seamed together to provide an overall virtual, macro image of the specimen at a lower resolution. Several original microscope views at higher magnifications are digitized and stored as digitized images coherently seamed together to provide virtual micro images at higher resolution. A data structure is formed with these virtual macro and micro digitized images along with their mapping coordinates. Preferably, a generic viewing program is also provided in the data structure that allows remote users to manipulate and interpret the tiled images on the user's monitor.

Abstract: This invention relates to an image processing method of quantizing a multivalued image to perform image processing. The luminance frequency of the multivalued image is calculated, a quantization threshold for quantization is specified on the basis of the calculated luminance frequency, a representative value used for quantization of the multivalued image is calculated on the basis of the specified quantization threshold and the luminance frequency, and the multivalued image is quantized using the calculated representative value. The quantization threshold is an average luminance value obtained when the histogram distribution converges to make the skew of the histogram distribution of the luminance frequency fall within a predetermined range. The representative value is an average luminance value in each distribution region of the histogram distribution of the luminance frequency that is divided by the quantization threshold.

Abstract: A method of detecting the presence and position of labelled material in a sample in which the labelled material either gives off light or can be stimulated to do so. The sample is imaged onto a CCD array which is scanned following each exposure. A cooled CCD camera or an image intensified CCD camera may be used. Measurements are performed on the data signals so obtained, to identify clusters of data values from adjacent regions of the CCD array caused by light incident on those regions. The measured signal values are compared with at least one threshold so as to distinguish clusters resulting from light emitting regions of labelled material from the remainder of the sample and the centroid of each light produced cluster of data values is computed with reference to the camera array, and a signal value corresponding to the centroid coordinates is stored in a memory together with the centroid coordinates of any other light produced clusters identified during the same interrogation.

Abstract: An image signal read out from an original is binary-coded on the basis of a binary-coding threshold value. The binary-coding threshold value is calculated by making a histogram representing a frequency distribution of tones of the image signal, calculating the value of a peak tone at which the frequency is maximized, calculating as a first tone at which the frequency becomes equal to a predetermined level on the lower tone side of the peak tone, calculating a reference binary-coding threshold value by adding to the peak tone the difference between the first tone and the peak tone, and determining the binary-coding threshold value on the basis of the reference binary-coding threshold value.

Abstract: A histogram equalization apparatus for contrast enhancement of a moving image includes a CDF calculator for counting the number of pixels having a gray level from the minimum to the maximum or less with respect to an input image in a frame unit to calculate a cumulative distribution function (CDF) value of each gray level, and a look-up table for updating histogram-equalized level corresponding to each gray level of the input moving image in a frame unit based on the CDF value of each gray level, and outputting a corresponding histogram-equalized level according to a level of the input moving image. Therefore, the histogram equalization apparatus can perform real-time histogram equalization on a moving image with a simple hardware, without using a frame memory, a probability density function (PDF) calculator and dividers for a cumulative density function (CDF) calculation.

Abstract: An image information recognition system is provided which quantifies an angle of rotation and/or magnification of a wiring pattern in an image to facilitate a subsequent image retrieval.

Abstract: An image is divided into blocks and coded by matching each block with a pattern having either a single region, or two or more regions defined by internal edges. The coded value of the block includes a pattern identifier, and a single pixel value for each region in the matching pattern. The coded image is decoded by constructing blocks divided into regions according to the pattern identifiers, and filling each region with the given single pixel value. The decoded blocks may be the same size as the blocks in the original image, or smaller, in which case the decoded image is reduced, or larger, in which case the decoded image is enlarged.

Abstract: An image desired to be reproduced is scanned to determine its video pixel grey values. A histogram generator generates a histogram distribution representing a frequency of the grey values. The histogram distribution is analyzed to determine a minimum and maximum input grey values which define input boundaries. A segment point is computed between the input boundaries based on the histogram data. The segment point defines a plurality of input segments between the input boundaries. A dynamic output range is selected. Each input segment is mapped to an output segment based on a linear transformation for the corresponding segment. In this manner, a tone reproduction curve map having a piecewise linear transformation is automatically generated from the image histogram data.

Abstract: A method and system implements a high addressability characteristic into an error diffusion process. A histogram of the image is generated and the actual background and black reference values are determined. A gray level value representing a pixel is received. The gray level value has a first resolution which corresponds to an original input resolution. The gray level value is interpolated to generate subpixel gray level values which correspond to a second resolution. The second resolution is higher than the first resolution and corresponds to the high addressability characteristic. A threshold circuit thresholds the interpolated gray level value and generates an error value as a result of the threshold using the determined background and black reference values. The error value has a resolution corresponding to the first resolution. A portion of the error value is diffused to adjacent pixels on a next scanline.

Abstract: In a digital signal processing system, a method for selecting a transform function to apply to an input signal based on characteristics of the signal, and for self-adjusting criteria which are used in selecting a transform function to apply to a subsequent signal. Characteristics are obtained from the signal. The characteristics are compared to adjustable criteria which are used in selecting a transform function. Differing criteria are maintained for the different selectable transform functions. A record is maintained of transform functions selected and the particular characteristics that caused the selection. Based on the ability of a transform function to minimally define the coded signal, an inverse transform function is selected to decode the signal. The criteria used in selecting a transform function to apply to a subsequent signal are adjusted based on a quality measure of the decoded signal and the record of selected transform functions.

Abstract: Compensation data 10 distilled from image sample population volume 12 compensates pixel records 25R and 25L of related images 26R and 26L. Each pixel address stores a parameter code defining a parameter level for that pixel along a code scale of levels of a parameter such as greyscale common to both related images. The pixel records are enhanced through compensation of these parameter codes to provide enhanced disparity information. Pixel pairer 28P establishes a set of pixel pairs from the related pixel records in image correspondence. Sample generator 28G determines discrepancy between the code levels of each pixel pair to provide image population 12 of code discrepancy samples between the pixel records. Each pixel pair established from the two pixel records yields a difference sample in the image population. Profile extractor 28E extracts population profiles 12E of code discrepancy samples from image population 12. Each population profile is a vertical slice (see FIG.

Abstract: A pattern inspection method capable of providing an improved inspection reliability includes the following steps. A pattern classification is set every pixel-value range of a reference-image data obtained from a reference pattern. Each of the pattern classifications has reference-data preparing parameters. With respect to each pixel of the reference-image data, pixel values of a pixel window which is composed of the pixel as a center pixel and neighbor pixels around the center pixel are checked to prepare a reference data. When all of the pixels in the pixel window are within a single pixel-value range, the center pixel is decided as a pattern uniform portion, and the reference data is prepared according to the parameters of a corresponding pattern classification.

Abstract: An image desired to be reproduced is scanned to determine its video pixel gray values. A histogram generator generates a histogram distribution representing a frequency of the gray values. The histogram distribution is analyzed to determine a background peak gray value of the image and a standard deviation of the histogram distribution based on a Gaussian approximation. A thresholding circuit dynamically adjusts the background peak value based on the standard deviation and a selected scaling factor to generate a background threshold value. The background threshold value expands a range of background gray values in the image which are eliminated during image reproduction. Eliminating substantially all background gray values improves the quality of the reproduced image.

Abstract: A method for automatically identifying the range of useful digital values to be displayed on a display medium and providing an appropriate gray scale transfer to optimize the diagnostic value of the final displayed image. One or more gray scale transfer functions corresponding to desired display media, one or more sets of experimentally determined constants, and one or more sets of algorithms are stored in a computer memory. A smoothed histogram and its histogram integral are constructed representing the frequency of occurrence of the digital values stored in the data bank. A low point and an edge point are identified from the histogram and integral, and based on the type of radiographic examination selected, are used with appropriate constants and algorithms to calculate a maximum and minimum value. Values lower than the minimum are replaced with the minimum, and values higher than the maximum are replaced by the maximum.

Abstract: Pixel data representative of a discrete pixel image is processed to identify structures and non-structures in the image defined by the pixel data. The structures and non-structures are processed in different manners. The structures are identified by computing gradient information on each pixel and by comparing the gradient information to a gradient threshold, and by comparing gradient directions for adjacent pixels to one another. The edges defining the structures are binary rank order filtered. The structures are orientation smoothed and sharpened. The non-structures are homogenization smoothed and original texture is blended back into non-structural regions.

Abstract: A method and apparatus for image classification includes a first embodiment, in which halftone detection is performed based on the size of a boundary set between a class of light pixels and a class of dark pixels, and further based upon image information contained within each single image plane (i.e within one color plane). This embodiment of the invention is based upon the distinctive property of images that halftone areas within the image have larger boundary sets than non-halftone areas within the image. A second, equally preferred embodiment of the invention provides a cross color difference correlation technique that is used to detect halftone pixels.

Abstract: Structures are identified in pixel data representative of a discrete pixel image and are differentiated from non-structures. The structures are identified by comparing gradient magnitudes with a selected gradient value and by comparing gradient directions of adjacent pixels. Pixels having desired relationships are counted and a focus parameter is applied to the resulting count. A final gradient threshold value is identified based upon the count and the focus parameter. The threshold value identifies structural pixels. Isolated adjacent regions may be eliminated from the structures. Edges defined by the structures are binary rank order filtered to arrive at the final structural definition. The structural and non-structural pixels may be further processed to enhance the image for subsequent display.

Abstract: A computer-aided diagnosis (CAD) method for the automated detection of lung nodules in a digital chest image, a computer programmed to implement the method, and a storage medium which stores a program for implementing the method, wherein nodule candidates are first automatically selected by thresholding the difference image and then classified in six groups. A large number of false positives are eliminated by adaptive rule-based tests applied to the original chest image and in the difference image and an artificial neural network (ANN) applied to remaining candidate nodule locations in the original chest image. Using two hundred PA chest radiographs, 100 normal and 100 abnormal, as the database, the presence of nodules in the 100 abnormal cases was confirmed by two experienced radiologists on the basis of CT scans or radiographic follow-up. The CAD method achieves, on average, the sensitivity of 70% at 1.7 false positives per chest image.

Abstract: A method and apparatus are provided to allow threshold values needed for the desired number of gradations to be set through a simple process when converting image data with numerous values into a low number of gradations. The desired number of gradations is achieved without shading being noticeable even for images having shading. The cumulative frequency distribution of the pixels for each brightness is determined from image data having a brightness range with designated gradations. The average brightness and standard deviation of the portion that is primarily background is determined from this cumulative frequency distribution. The first threshold value is determined on the basis of this average brightness and standard deviation, the second through n-1st threshold values corresponding to the meaningful information are determined on the basis of this first threshold value. An n value conversion process is accomplished using these threshold values.

Abstract: An apparatus and method for magnifying a dynamic range of an image while improving a contrast of an image are provided. A histogram calculator receives a digital image signal and calculates histograms depending on the illuminance distribution of an image. A histogram accumulator receives a histogram output by the histogram calculator, obtains an accumulative density function by integrating the input histogram up to each illuminance level and normalizes the accumulative density function, and generates an initial conversion function. A bias calculator determines a threshold value for dividing into a low illuminance level and a high illuminance level on the basis of a histogram output by the histogram calculator, and obtains a histogram compensation function for applying a negative bias to a histogram around the threshold value.

Abstract: A method for automatically forming a filtered subset of digital values from a first set of digital values corresponding to a plurality of picture elements representing a radiographic exposure of a subject for use in displaying a radiograph of said subject, the method comprising: (1) generating a first histogram of the first set of digital values; (2) generating an integral curve representing an integral of the first histogram; (3) identifying an inflection point and corresponding digital value V.sub.peak1 on the first histogram integral where the integral curve first changes slope in a direction decreasing from the highest digital value; (4) calculating a parameter V.sub.test =V.sub.

Abstract: Analog image data a SEM are converted into digital data, and are processed by a spatial filtering processing, histogram processing, threshold value setting, three-valued image data processing, noise reduction and the like. Area of a pattern in the three-valued image data is calculated by a labelling and calculation processing, and a pattern is sequentially detected by comparing the area of the pattern with a reference area value. The comparison and detection of the same or similar patterns repeated in the SEM image are performed by using the area of the pattern, and are not performed by a shape of the pattern, thereby resulting a precise detection at high speed by using a microprocessor. Since it is possible to perform a pattern recognition from the area value even though the pattern does not have a characteristic, it is possible to precisely detect and recognize a pattern image in high speed.

Abstract: Foreground information is separated from background information on a document where the foreground information is superimposed in an additive or subtractive manner on the background information. The document is divided into a plurality of segments and a frequency distribution of extreme values of the signal intensity in each of the segments is determined. The frequency distribution so determined exhibits one main mode (background mode) which essentially contains the background information. If an adequate foreground/background separation cannot be achieved with the first two steps, the determination of a frequency distribution of signal intensities of the whole document is effected in a third step and a sharper separation effected by a fourth step in which a joint evaluation of the two frequency distributions is determined.

Abstract: A multi-tone-level image is converted into an M-tone-level image using M-1 threshold values, where M-1.gtoreq.2. Middle-tone-level pixels in the M-tone-level image are treated as unknown pixels, tone levels other than the middle tone levels are given to the unknown pixels based on the states of pixels in the neighborhood of the unknown pixels, and thus produces an N-tone-level image, where M>N.gtoreq.2. Two threshold values of the M-1 threshold values are used for determine the middle-tone-level pixels, the two threshold values corresponding to the lightest tone level of a first image region and the darkest tone level of a second image region respectively. The first image region has tone levels which are darker than the middle tone levels and the second image region has tone levels which are lighter than the middle tone levels.

Abstract: The present invention is a hybrid technique for finding defects on digitized device images using a combination of spatial domain and frequency domain techniques. The two dimensional spectra of two images are found using Fourier like transforms. Any strong harmonics in the spectra are removed, using the same spectral filter on both spectra. The images are then aligned, transformed back to the spatial domain, and subtracted. The resulting spectrally-filtered difference image is thresholded and analyzed for defects. Use of the hybrid technique of the present invention to process digitized images results in the highest-performance and most flexible defect detection system. It is the best performer on both array and random devices, and it can cope with problems such as shading variations and the dark-bright problem that no other technique can address.

Abstract: A red-eye reduction system is described that includes a masking module. The masking module converts an image into a mask having first state areas representing red color pixels of the image and second state areas representing other color pixels of the image. The image includes an eye with a red pupil. A pupil locating module is coupled to the masking module to locate a substantially first state area in the mask that resembles a pupil. A color replacing module is then coupled to the pupil locating module to change the red color pixels in the area into monochrome (grey) or other predefined colors. The color replacing module also adjusts the boundary of the area by changing the colors of pixels in close proximity to the area if the color of these pixels is determined to be sufficiently close to red such that natural appearance of the eye is maintained when reducing the red pupil. A method of reducing red-eye effect in a digital image is also described.

Abstract: In a digital signal processing system, a method for selecting a transform function to apply to an input signal based on characteristics of the signal, and for self-adjusting criteria which are used in selecting a transform function to apply to a subsequent signal. Characteristics are obtained from the signal. The characteristics are compared to adjustable criteria which are used in selecting a transform function. Differing criteria are maintained for the different selectable transform functions. A record is maintained of transform functions selected and the particular characteristics that caused the selection. Based on the ability of a transform function to minimally define the coded signal, an inverse transform function is selected to decode the signal. The criteria used in selecting a transform function to apply to a subsequent signal are adjusted based on a quality measure of the decoded signal and the record of selected transform functions.

Abstract: The image correlation method and apparatus correlates or matches a test image with a template. The image correlation apparatus includes an image processor for partitioning the template into a number of labels, for determining the total number of pixels N.sub.T which form the template and for determining the number of pixels N.sub.i which form each of the labels i. The image correlation apparatus also includes comparison means for comparing the test image to the template. The comparison means determines, for each predetermined gray level j, the number of pixels of the test image N.sub.j,i representative of a predetermined gray level j which correspond to a predetermined label i of the template. The comparison means also determines, for each predetermined gray level j, the number of pixels of the test image N.sub.j representative of a predetermined gray level j which correspond to the template.

Abstract: A method for automatically locating a desired region of interest (ROI) in an original digital image involving the use of histograms to determine a threshold that is used to generate a mask to detect the ROIs. Digital image intensity data is obtained and a histogram of the intensity data is formed. The threshold is determined by first determining a histogram point that represents the noise peak in the image intensity data and then setting the threshold to be the image intensity value following the noise peak that first exhibits a positive curvature. The threshold is used to create a mask from the original digital image, and contiguous areas of pixels are defined in the mask. At least one of these contiguous areas is determined to be a desired ROI.

Abstract: Image analysis methods and systems are used for identifying objects in a background by defining a data space, such as a histogram or a color space. The data space comprises a plurality of sub-spaces, which could be selected based, for example, on a histogram, or on the way pixel values or color parameters cluster in their respective spaces. The threshold values in the data space are selected, a list of all ordered pairs of thresholds is generated to define multiple data sub-spaces, and the image is multiply searched, once in each sub-space for at least one representation of a candidate object, where the candidate object has at least one predetermined attribute value. Valid objects are identified by comparing the candidate object attribute values to a defined set of valid object attribute values contained in a driver.

Abstract: A method and system for the computerized registration of radionuclide images with radiographic images, including generating image data from radiographic and radionuclide images of the thorax. Techniques include contouring the lung regions in each type of chest image, scaling and registration of the contours based on location of lung apices, and superimposition after appropriate shifting of the images. Specific applications are given for the automated registration of radionuclide lungs scans with chest radiographs. The method in the example given yields a system that spatially registers and correlates digitized chest radiographs with V/Q scans in order to correlate V/Q functional information with the greater structural detail of chest radiographs. Final output could be the computer-determined contours from each type of image superimposed on any of the original images, or superimposition of the radionuclide image data, which contains high activity, onto the radiographic chest image.