Abstract: The invention provides a reproducible, objective quantification technique that reliably segments white matter structures. The technique receives a seed voxel within the white matter structure from an individual, determines thresholds and selection criteria, creates a binary mask based on the at least one threshold and the at least one selection criteria and calculates the boundary of the white matter structure based on the binary mask. A magnification factor is applied to each component of the eigenvectors of voxels. Boundary voxels are determined wherein each of the boundary voxels has a magnitude above a predetermined value and is located next to a voxel having a magnitude below the predetermined value. A vector is drawn from the seed voxel to a boundary voxel and the boundary voxels are connected together, thereby forming the region of interest within the connected boundary voxels.

Abstract: Certain embodiments of the present invention provide systems and methods for correlating an acquired image with an historical image. Certain embodiments include registering an acquired image and at least one historical image according to a coordinate system, applying a metric to the acquired image and each of the historical image(s), and identifying a correlation between the acquired image and one of the historical image(s) based on the metric. Certain embodiments further include storing the correlation between the acquired image and an historical image for automatic linking of the acquired image and the historical image upon display. In certain embodiments, the acquired image and the historical image may be displayed based on the correlation. In certain embodiments, the acquired image and the historical image may be automatically linked based on the correlation. In certain embodiments, the metric analyzes the acquired image and at least one historical image based on anatomy.

Abstract: Certain embodiments of the present invention provide a method for improving workflow in assessing lesions including displaying a reference image and determining a lesion contour. The reference image is displayed with a user interface component. The lesion contour is determined with a contour processing component. The lesion contour is based at least in part on the reference image and a threshold value.

Abstract: Methods and systems for generating computed tomographic (CT) images from image data acquired during different biological cycles are provided. A computer is programmed to receive a plurality of scan data acquired during a gated acquisition window of each of a plurality of biological cycles, blend the scan data acquired during a first of the plurality of biological cycles with the scan data acquired during a second of the plurality of biological cycles, and construct a final image from the blended data.

Abstract: A method for navigating digital medical images includes providing a digitized patient medical image of a structure of interest in a patient, using a pre-trained classifier to segment the structure of interest from the image, creating links from the structure of interest to a corresponding structure in an anatomical atlas, receiving a query to view the structure of interest, parsing the query to identify one or more keywords from noun phrases in the query, mapping a keyword to a corresponding structure in the anatomical atlas, wherein the anatomical atlas structure is associated with a link to the corresponding structure in the patient image, and following the link to display said patient structure of interest.

Abstract: A system and method for vertebrae and intervertebral disc localization in magnetic resonance (MR) images is disclosed. In this method, a target spine MR image is segmented by generating a binary segmentation. A spinal cord line is then estimated on the segmented image. Neighboring vertebrae in a region of interest around the estimated spinal cord line are then localized by labeling vertebral regions based on characteristics, such as size and shape, of regions of connected segmented pixels within the regions of interest. A center point of an intervertebral disc between the neighboring vertebrae is then calculated by finding a point halfway between respective borders of the vertebral regions on a line connecting the centers of mass of the vertebral regions.

Abstract: A branch extension method and system for segmenting airways in 3D image data is disclosed. An initial airway segmentation is obtained from the 3D image data. Terminal branches of segmented airways of the initial airway segmentation are identified. The segmentation of the terminal branches is then extended. The segmentation of the terminal branches can be extended using various segmentation techniques. This method can use complex segmentation techniques to extend the terminal branches without having a large impact to the overall speed of the segmentation.

Abstract: A tomographic reconstruction method or device for calculation of a 3-dimensional image data or 2-dimensional slice from a limited amount of 2-dimensional or 1-dimensional projection images using iterations to calculate 3D image elements, comprising transformation of measured data to transform domain.

Abstract: Disclosed is a systematic way of automatically segmenting lung regions. To increase the efficiency of a lung segmentation technique, a region-based technique, such as region growing, is performed by a computer on a middle slice of the CT volume. A contour-based technique is then used for a plurality of non-middle slices of the CT volume. This allows the implementation to be multithreaded and results in an improvement in the segmentation algorithm's efficiency.

Abstract: An adaptive density correction (ADC) method and system automatically compensate for pseudo-enhancement (PEH) of voxels in computed tomography (CT) data, such as in fecal-tagged CT colonography (ftCTC), so air (or another low-contrast background) and soft tissues are represented by their usual CT attenuations. ADC estimates an amount of pseudo-enhancement energy that was received by voxels that are near tagged voxels (i.e., voxels that are tagged with a high-contrast agent), based on a first distribution scheme, such as a Gaussian distribution. ADC then iteratively distributes PEH energy received by voxels to neighboring voxels, according to another distribution scheme, which may be another Gaussian function. ADC then subtracts the total amount of PEH energy at each voxel from the CT data of the voxel.

Abstract: An adaptive density mapping (ADM) method and system automatically identify interface regions between air and material tagged with contrast agents in computed tomographic (CT) image data, then map CT attenuations of voxels outside the identified interface regions, such that voxels that represent tagged material are made to represent air or another gas.

Abstract: A method of processing checks includes obtaining first information relating to the checks that is compiled by an operator using a calculator, such as by scanning a calculator tape, wherein the first information includes a first number of first dollar amounts, each of which corresponds to a respective one of the checks. The method further includes scanning the checks to create a plurality of check images, obtaining from the check images a second number of second dollar amounts, each corresponding to a respective one of the check images, determining whether the first number is equal to the second number, and providing an error indication if the first number does not equal the second number (which is an indication that a misfeed has occurred).

Abstract: A pattern inspection apparatus uses a die-to-database comparison method which compares detected pattern data obtained from an optical image of a pattern of a plate to be inspected with first reference pattern data obtained from designed pattern data in combination with a die-to-die comparison method which compares the detected pattern data with second reference pattern data obtained by detecting an area to be a basis for repetition. A computer detects presence of a plurality of repeated pattern areas from layout information contained in the designed pattern data, reads the arrangement, the number, the dimension and the repeated pitch of the repeated pattern areas, and automatically fetches an inspection area of the die-to-die comparison method.

Abstract: The object of the invention is to provide a defect inspection method, a defect inspection system, and a defect inspection program, in which the inspection time for the object to be inspected and identifiable as a linear or band-like one is so very short that the operation time and tact time can be cut down, and which prevents misidentification even when there is distortion or deformation in the object, thereby making precise defect inspection possible. In the defect inspection method, system and program, the luminance and the number of pixels in the widthwise direction of the image of the object to be inspected such as the edge portion of a semiconductor wafer, wherein the image is identifiable as a linear or band-like one, are summated up, and the object is determined as defective when the value of the summation is below a reference value.

Abstract: A method and apparatus for identifying 3-D coordinates of a target region on a tire includes: taking a digital image of a tire; finding an edge of a tire bead using pixel brightness values from the tire image; calculating tire bead circle center and radius using a plurality of image pixels on the tire bead edge; and performing a pixel brightness search around the bead circumference using the bead circle's center and radius to identify the target area X, Y coordinates. The Z-coordinate and slope of the target area are determined from multiple point distance calculations across the region.

Abstract: There is disclosed a method apparatus for image processing of e.g. MRI images. A method and apparatus is disclosed enabling one to determine if an image contrast/intensity change of a chosen point within an imaged subject volume (e.g. a lesion) is an artefact of motion in the imaged subject. An apparatus and method for fast visual assessment of the imaged subject volume's alignment quality (within a time-sequence of images) is provided, and an interactive display tool enabling effective use of this method is provided.

Abstract: A system and method for automatically modeling a three dimensional object, such as a face, from a single image. The system and method according to the invention can construct one or more three dimensional (3D) face models using a single image. Although use of a frontal image simplifies the processing, the system and method according to the invention can also generate a 3D model from a non-frontal image (e.g., an image where the object to be modeled has an out of plane rotation, such a face that is turned to one side to some extent). The system is fully automatic. It is fast compared to the other modeling systems, Furthermore, it is very robust. It can be used to construct personalized models for games, on-line chat, and so on. It can also be used as a tool to generate a database of faces with various poses which are needed to train most face recognition systems.

Abstract: Systems and methods are provided for generating training samples for a classification system associated with a vehicle occupant protection system. A three-dimensional virtual reality scene is generated (138) from a cockpit model having a set of at least two user configurable parameters (132) and an occupant model having a set of at least two user configurable parameters and a plurality of points of articulation (134). A plurality of training samples are generated from the virtual model (142). Each of the plurality of training samples has a first set of parameters associated with the cockpit model, a second set of parameters associated with the occupant model, and an occupant pose defined according to the points of articulation associated with the occupant model.

Abstract: An alpha image encoding and decoding scheme is disclosed. In the encoding an alpha image, is decomposed into image blocks (600) comprising multiple image element (610). The blocks (600) are compressed into block representations (700). A block representation (700) comprises at least a color codeword (710), an alpha codeword (720), an alpha modifying codeword (730) and a sequence (740) of alpha modifier indices. The color (710) and alpha (720) codeword (710) are representations of the colors and alpha value of the image elements (610) of the block (600), respectively. The alpha modifying codeword (730) is a representation of a set of multiple alpha modifiers for modifying an alpha value represented by the alpha codeword (720). The index sequence (740) includes an alpha index for each image element (610) in the block (600), where an alpha index identifies one of alpha modifiers in the alpha modifier set.

Abstract: A content-based approach is provided for detecting near duplicate images. The approach generally involves analyzing the content of images to be compared and generating color and texture histogram descriptor data for each image. The images may then be compared based upon the color and texture histogram descriptor data to determine whether the images are near duplicate images. Content-based image signature data may also be generated for each of the images based upon the color and texture histogram descriptor data. The image signature data may then be compared to determine whether the corresponding images are near duplicate images.

Abstract: A method for histogram calculation using a graphics processing unit (GPU), comprises storing image data in a two-dimensional (2D) texture domain; subdividing the domain into independent regions or tiles; calculating in parallel, in a GPU, a plurality of tile histograms, one for each tile; and summing up in parallel, in the GPU, the tile histograms so as to derive a final image histogram.

Abstract: A text detection technique comprises local ramp detection, identification of intensity troughs (candidate text strokes), determination of stroke width, preliminary detection of text based on contrast and stroke width, and a consistency check.

Abstract: The invention relates to a system and method for recognition, classification and spatial localization of bounded 3D-objects. In particular it relates to a computerized method for recognition, classification and localization of objects. The method comprises generation of a training database based on a large number of training views recorded by a camera or constructed using a CAD representation of an object. Characteristic curves are derived from the training views, and primitives of the curves are detected. Intrinsic and extrinsic descriptors of features are stored in the database together with data about the object class and pose of the view. Finally the recognition takes place in two stages: First the intrinsic descriptors of the recognition view are compared with those of the database. Second, among the best matching features it is explored which features agree mutually in the sense that they suggest the same object class at the same pose.

Abstract: A device includes a first change detection unit which detects a change region distribution based on the texture information; a second change detection unit which detects a change region distribution based on a background differential method; a texture detection unit, a change-detection-unit weighting storage unit, and a change detection result integration unit, wherein while applying a firm texture differential method to the illumination change as much as possible, a detection unit based on a background differential method is applied to a region to which the texture differential method is not favorably applicable.

Abstract: The present invention utilizes generic and user-specific features of handwriting samples to provide adaptive handwriting recognition with a minimum level of user-specific enrollment data. By allowing generic and user-specific classifiers to facilitate in a recognition process, the features of a specific user's handwriting can be exploited to quickly ascertain characteristics of handwriting characters not yet entered by the user. Thus, new characters can be recognized without requiring a user to first enter that character as enrollment or “training” data. In one instance of the present invention, processing of generic features is accomplished by a generic classifier trained on multiple users. In another instance of the present invention, a user-specific classifier is employed to modify a generic classifier's classification as required to provide user-specific handwriting recognition.

Abstract: A method and apparatus for detecting and locating objects of interest in video sequences is provided. A frame is defined as an image belonging to video sequences. Each frame with the same or different size of original input sequences is searched by the same or different size window efficiently for detecting objects. The characteristics of temporal redundancies in video sequences are used in detecting objects in video sequences.

Abstract: The present invention provides a technique for retrieving pictures from a large database that is less complex and uses significantly less memory and computational resources than current techniques. This is accomplished by determining representative data vectors based on a tolerance distance that represents data vectors in a given vector space that defines the picture to extract features of the picture that facilitates in retrieving pictures.

Abstract: An image processing apparatus is made up of a processing target area specification section for specifying a processing target area; a circumferential direction specification section for specifying the circumferential direction of a contour circle in the processing target area; an edge detection region specification section for specifying three or more regions made different in position in the circumferential direction in the processing target area as edge detection regions based on the processing target area and the circumferential direction; an edge position identification section for identifying edge positions with respect to the radial direction based on the intensity distribution in the edge detection regions; and a contour circle determination section for determining a contour circle based on the edge positions identified relative to the edge detection regions.

Abstract: Method and system for comparing two images using a computing system is provided. The method comprises comparing a first image with a second image, the first image and the second image described by one or more image attributes. The images further comprise a plurality of components described by one or more component attributes. The comparison of first image with second image is performed by comparing components of first image with components of the second image using a hierarchy of component attributes. A list of matched and unmatched components between the first image and second image is created. A report is generated out of the list based on report options.

Abstract: A method and system for discovering from a database an object which is confusingly similar with a known object. A database, such as the internet, is searched for objects which, when discovered, may be duplicated and stored. A determination is then made if any object from the database is confusingly similar with a known object. A computer-readable storage medium storing program for causing a processing system to perform the steps of duplicating objects from a database to produce duplicated objects; storing the duplicated produce to produce stored duplicated objects; and determining if any stored duplicated object is confusingly similar with a known object.

Abstract: Systems and methods are provided for classifying an input feature vector, representing a vehicle occupant, into one of a plurality of occupant classes. A database (106) contains a plurality of feature vectors in a multidimensional feature space. Each feature vector has an associated class from the plurality of output classes. A data pruner (108) eliminates redundant feature vectors from the database. A data modeler (109) constructs an instance-based, non-parametric classification model (110) in the multidimensional feature space from the plurality of feature vectors. A class discriminator (112) selects an occupant class from the plurality of occupant classes according to the constructed classification model.

Abstract: A method constructs a strong classifier from weak classifiers by combining the weak classifiers to form a set of combinations of the weak classifiers. Each combination of weak classifiers is boosted to determine a weighted score for each combination of weak classifiers, and combinations of weak classifiers having a weighted score greater than a predetermined threshold are selected to form the strong classifier.

Abstract: This invention switches the number of bits per pixel between object types when converting code information into image data, thereby rendering one entire page at high speed while minimizing the influence on print quality. For this purpose, an apparatus analyzes received PDL data, and generates a DL (Display List) serving as a rendering unit. The apparatus determines whether the DL is a character/line image or a halftone image including a photograph image. When determining that the DL is a character/line image, the apparatus resizes the DL in accordance with the resolution of a printer engine, and renders each color component at an 8-bit precision (integer). When determining that the DL is a halftone image including a photograph image, the apparatus resizes the DL in accordance with the resolution of the printer engine, and renders each color component at a 10-bit precision (8-bit integral part+2-bit decimal part).

Abstract: A motion-vector detector determines the centroid of pixels on a reference frame that is identified with position information set in a database and associated with a feature address corresponding to a feature of a target pixel. The motion-vector detector detects, as a motion vector of the target pixel, a vector that has a starting point at a pixel on the reference frame which corresponds to the target pixel on a current frame and has an end point at the determined centroid. The present invention can be applied to an apparatus for generating a motion vector and allows prompt detection of a motion vector.

Abstract: A motion-vector detector determines the centroid of pixels on a reference frame that is identified with position information set in a database and associated with a feature address corresponding to a feature of a target pixel. The motion-vector detector detects, as a motion vector of the target pixel, a vector that has a starting point at a pixel on the reference frame which corresponds to the target pixel on a current frame and has an end point at the determined centroid. The present invention can be applied to an apparatus for generating a motion vector and allows prompt detection of a motion vector.

Abstract: An overall motion detecting device (1) includes a search region variable-shaped motion detector (2) and an overall motion information generator (3). The overall motion information generator (3) controls a surveillance camera (5) which is an external image capturing system and supplies control information of the surveillance camera (5) (such as moving direction, angular speed and focal length) to the search region variable-shaped motion detector (2) as overall motion information (S3). The search region variable-shaped motion detector (2) calculates a motion vector (mv) while renewing, by picture, a search region having a search shape determined by the overall motion information (S3) and a region area suitable for its operational capability.

Abstract: An overall motion detecting device (1) includes a search region variable-shaped motion detector (2) and an overall motion information generator (3). The overall motion information generator (3) controls a surveillance camera (5) which is an external image capturing system and supplies control information of the surveillance camera (5) (such as moving direction, angular speed and focal length) to the search region variable-shaped motion detector (2) as overall motion information (S3). The search region variable-shaped motion detector (2) calculates a motion vector (mv) while renewing, by picture, a search region having a search shape determined by the overall motion information (S3) and a region area suitable for its operational capability.

Abstract: An imaging system incorporating adaptive compression which includes determining linear predictive differential residuals from an imager array pixel row. The differential residuals are classified into categories, each category having a range of differential residuals associated with it. The categories are analyzed to produce an ordered list having categories with most to least frequent residuals falling within a respective residual range associated with a respective category. The ordered list is then used to select a variable length encoding table with a matching ordered list. Variable length encoded category and range position offset data is output to a serializer unit, where the range position offset refers to a position in a range associated with a particular category.

Abstract: A device for controlling the representation of information on a vehicle screen includes an actuator which can be rotated about a longitudinal axis and deflected transversely with respect to the longitudinal axis. The portion (2) of the information which is represented on the vehicle screen displaced by a deflection movement of the actuator transversely with respect to the longitudinal axis (z), and discrete markable elements can be marked by a rotational movement of the actuator about the longitudinal axis. The information includes the discrete markable elements.

Abstract: A method and system is provided for automatically generating metadata for a digital image based on ambient conditions. Aspects of the preferred embodiment include determining a plurality of ambient conditions associated with a location of a subject of a digital image captured with a mobile image capture device, the ambient conditions existing at a time related to a time of capture of the digital image; aggregating the plurality of ambient conditions into at least one ambient environment descriptor based on user-definable aggregation rules; and associating the ambient environment descriptor with the digital image as metadata.

Abstract: A directed pattern enhancement method receives a learning image and pattern enhancement directive. Pattern enhancement learning is performed using the learning image and the pattern enhancement directive to generate pattern enhancement recipe. An application image is received and a pattern enhancement application is performed using the application image and the pattern enhancement recipe to generate pattern enhanced image. A recognition thresholding is performed using the pattern enhanced image to generate recognition result. The pattern enhancement directive consists of background directive, patterns to enhance directive, and patterns to suppress directive. A partitioned modeling method receives an image region and performs feature extraction on the image region to generate characterization feature. A hierarchical partitioning is performed using the characterization feature to generate hierarchical partitions. A model generation is performed using the hierarchical partitions to generate partition model.

Abstract: A digital image processing system (134) receives a signal S(x,y,t,c) where (x,y) are pixel coordinates, t is time, and c is a color component. The system generates the signal S*(x,y,t,c) as a function of a linear combination of the values S(x?,y?,t?,c) over a set A of points (x?,y?,t?) with plural values (x?,y?) and plural values t?, wherein each value S(x?,y?,t?,c) in the linear combination is multiplied by a coefficient which is a function of the difference S(x,y,t,c)?S(x?,y?,t?,c). In some embodiments, one or more of the values S(x?,y?,t?,c) are replaced with previously calculated values S*(x?,y?,t?,c) to save memory.

Abstract: A method of reducing noise in an image including steps for obtaining a first value for a target pixel, obtaining a respective second value for each neighboring pixel surrounding the target pixel and having the same color as the target pixel, for each neighboring pixel, comparing a difference between said first value and said second value to a threshold value, and replacing the first value with an average value obtained from the first value and at all second values from the neighboring pixels which have an associated difference which is less than or equal to the threshold value based on a result of the comparing step.

Abstract: An image processing apparatus includes a threshold selecting unit that selects a threshold based on an input gray-scale value of the target pixel. When a degree of flexibility is defined as representing a degree of the number of pixels that can take different output values as a result of threshold processing on the pixels with the same input gray-scale value, the threshold selecting unit selects a threshold from at least two types of thresholds associated with pixel positions and having the same period and different degrees of flexibility.

Abstract: An image processing device, method, recording medium, and program where the device includes a simple-type angle detecting unit simply detects the angle as continuity using correlation from an input image. The device includes an input configured to input image data made up of a plurality of pixels acquired by real world light signals being cast upon a plurality of detecting elements, and a real world estimating unit configured to estimate light signals being cast in an optical low-pass filter.

Abstract: A captured image of an object is input twice using a white background and a black background, after which the opacity D and the color image X of the object are obtained to yield an object image. It is possible to easily create a high-quality real image.

Abstract: A method for removing streaks from a scanned image includes performing a scanner light source OFF scan to generate a dark image having M rows of pixels and N+W columns of pixels; generating a line noise profile for the N columns of pixels of the dark image; performing a scanner light source ON scan to generate a regular image having M rows of pixels and N+W columns of pixels; processing the W columns of pixels of the regular image to detect a magnitude of the streak in the regular image; use the magnitude of the streak and the line noise profile to predict an amount of dark noise removal for each of the N columns of pixels of the regular image at the location of the streak; and subtracting the predicted amount of dark noise removal for each of the N columns of pixels from the respective N columns of pixels of the regular image.

Abstract: An image reading apparatus having a visible light source for emitting visible light, an invisible light source for emitting invisible light, an imaging optical system for irradiating a document by these light sources to form optical images of the document, and a CCD for photoelectrically converting the optical images of the document turns on the visible light source to acquire a visible light image signal by the CCD. Then, the apparatus turns on the invisible light source to acquire an invisible light image signal by the CCD. After the invisible light image signal is acquired, the apparatus turns off the invisible light source, and turns on the visible light source regardless of whether to read the document.

Abstract: A sample inspection system and method for displaying images of plural sample layers and a predetermined planar region within an image having a very large imaging region. Line sensors generate sample picture data for the plurality of sample layers, a computer stores the generated sample picture data, allows a planar region to be designated by a user for display, extracts picture data corresponding to the designated planar region for each of the respective layers from the sample picture data in response to designation of the planar region, and stores the extracted picture data as a set, and an image processing apparatus for displaying extracted pictures of the respective layers on a display in either an alternative way in which each picture may be selectively displayed individually and alternately, or simultaneously in which two or more of the pictures may be displayed together in a parallel or superposed manner.

Abstract: It is an object of the present invention to detect an image correction method from corrected images, in order to grasp a correction method of a correctly corrected images as a guideline for correction of a harmful image, and to present an image correction proposal to which a correction method in the past is applied, in order to indicate to a corrector a guideline for correction of the harmful image. In order to achieve the above-described objective, the present invention employs the following configuration. With respect to a harmful image data before correction and an image data after correction, a scene feature quantity representing the feature quantity of a scene for each scene of a video is calculated, and by comparing chronological lists of the respective scene feature quantities of the videos, a correction method for the scene configuration is detected.