Abstract: There is described a system and a method for generating a metatag indicative of an attribute of a person represented in an image, the method comprising: performing a computer-based analysis of the image; identifying a feature related to the attribute of the person in the image using results from the computer-based analysis; generating the metatag based on the identified feature; and associating the metatag to the image, the metatag being thereby related to the attribute of the person.

Abstract: A method for comparing a plurality of geometrical data representations each representing a spatial boundary surface of a corresponding geometrical object which surface changes over a selected extent of the object bounded thereby through providing the plurality of geometrical data representations on a common format basis, including scaling so as to each to have a common selected extent, to thereby result in a plurality of standardized spatial boundary surface geometrical data representations and comparing them at a plurality of matching section locations along each of the common extents at each of which there is a section outline curve representations. Comparing selected features of the commonly scaled section outline curve representations for such representations at corresponding ones of the selected matched section locations provides a basis for determining similarity therebetween.

Abstract: Methods and apparatuses process images. The method according to one embodiment accesses digital image data representing an image including an object; accesses reference data including a shape model relating to shape variation of objects from a baseline object, the objects and the baseline object being from a class of the object; and removes from the image an element not related to the object, by representing a shape of the object using the shape model.

Abstract: A system and method for tracking features, e.g., facial features, is provided, which allows for the tracking of features which move in a series of images and whose shape changes nonlinearly due to perspective projection and complex 3D movements. A training set of images is processed to produce clustered shape subspaces corresponding to the set of images, such that non-linear shape manifolds in the images are represented as piecewise, overlapping linear surfaces that are clustered according to similarities in perspectives. A landmark-based training algorithm (e.g., ASM) is applied to the clustered shape subspaces to train a model of the clustered shape subspaces and to create training data.

Abstract: A method for selecting vertices for performing deformable registration of imaged objects is provided. The selected vertices form corresponding pairs, each pair including a vertex from a first imaged object and a vertex from a second imaged object. The corresponding vertex pairs are sorted in order of distance between the vertices making up the corresponding vertex pair. The corresponding vertex pair with the greatest distance is given top priority. Corresponding vertex pairs that lie within a selected distance from the selected corresponding vertex pair are discarded. In this manner, the number of vertex pairs used for deformable registration of the imaged objects is reduced and therefore allows for processing times that are clinically acceptable.

Abstract: Methods and apparatus, including systems and computer program products, for creating high dynamic range images. One method includes actions of establishing a current set of digital image; displaying a high dynamic range image and information identifying the current set of images, the high dynamic range image being computed from the current set of images; and repeatedly, as long as the user provides user input changing the current set of images, receiving user input changing the current set of images and in response to each user input displaying a high dynamic range image and information identifying the current set of images, where the high dynamic range image is computed from the current set of images.

Abstract: A scene-based non-uniformity correction method super-resolution for eliminating fixed pattern noise in a video having a plurality of input images is disclosed, comprising the steps of warping each of the plurality of images with respect to a reference image to obtain a warped set of images; performing one of averaging and deblurring on the warped set of images to obtain an initial estimate of a reference true scene frame; warping the initial estimate of the reference true scene frame with respect to each of the plurality of images to obtain a set of estimated true signal images; performing a least square fit algorithm to estimate a gain image and an offset image given the set of estimated true signal images; applying the estimated gain image and estimated offset image to the plurality of images to obtain a clean set of images; and applying a super-resolution algorithm to the clean set of images to obtain a higher resolution version of the reference true scene frame.

Abstract: Methods and an apparatus process images. The method according to one embodiment accesses digital image data representing a plurality of objects with a plurality of features; generates a shape model for shapes of the plurality of objects, the step of generating a shape model including generating a baseline object by aligning the plurality of objects, and determining deformation modes to describe shape deformations between shapes of the plurality of objects and the baseline object; performs shape registration for the plurality of objects by fitting shapes of the plurality of objects using combinations of the deformation modes, to obtain registered shapes; and generates a probabilistic atlas by mapping the plurality of features from the plurality of objects to the baseline object, using the registered shapes for the plurality of objects.

Abstract: A method of producing an enhanced Active Appearance Model (AAM) by combining images of multiple resolutions is described herein. The method generally includes processing a plurality of images each having image landmarks and each image having an original resolution level. The images are down-sampled into multiple scales of reduced resolution levels. The AAM is trained for each image at each reduced resolution level, thereby creating a multi-resolution AAM. An enhancement technique is then used to refine the image landmarks for training the AAM at the original resolution level. The landmarks for training the AAM at each level of reduced resolution is obtained by scaling the landmarks used at the original resolution level by a ratio in accordance with the multiple scales.

Abstract: A method and system for detecting a source multifeatured three-dimensional object from two-dimensional source imagery, and for locating a best-matching three-dimensional object from a candidate database of such objects by performing a viewpoint and lighting invariant search among the candidate objects. The invention further includes the method and system for determining the optimally-fitting viewpoint, lighting, and deformation of the located three-dimensional candidate and basing an identification of the candidate with the source on the quality of fit between projections of the candidate and the source imagery.

Abstract: A method for adjusting an image projected onto a warped surface is provided. The method initiates with capturing calibration points within a calibration image projected onto the warped surface. The method includes analyzing the calibration points to determine relative locations among the calibration points. In one embodiment, the method utilizes a pseudo origin for locating beginning points for successive rows of calibration points. The calibration points are ordered according to the relative locations. The ordered calibration points may be applied to alter the image data prior to projection of the image data onto the warped surface. A system for calibrating points so that an image projected onto a warped surface does not appear as distorted is also provided.

Abstract: Disclosed is method of visual search for objects that include straight lines. A two-step process is used, which includes detecting straight line segments in an image. The lines are generally characterized by their length, midpoint location, and orientation. Hypotheses that a particular straight line segment belongs to a known object are generated and tested. The set of hypotheses is constrained by spatial relationships in the known objects. The speed and robustness of the method and apparatus disclosed makes it immediately applicable to many computer vision applications.

Abstract: An image processing apparatus has a distortion correction processing unit. The image processing apparatus has a distortion correcting range calculating unit that calculates an input image range necessary for distortion correction processing of the distortion correction processing unit. Thus, the corrected image (output image) obtained by the distortion correction processing can be outputted without overs and shorts for the image output range. The distortion is corrected by effectively using input (picked-up) image data, serving as the original data. Further, the distortion including the pincushion distortion, the barrel distortion, and the curvilinear distortion is effectively corrected.

Abstract: Systems and methods for detecting copies of documents are described. In one example, a system and method for detecting copies of documents utilizes multiple correlations of original and scanned copy detection patterns. The system captures a CDP that may be skewed or improperly sized. The system also retrieves or reconstitutes the expected CDP. In order to provide a more reliable comparison and without manipulating the captured CDP, the system manipulates the orientation or rotation and resolution or size of the retrieved or reconstituted CDP to register it with the captured CDP. Then the system performs a first correlation to determine if the captured CDP indicates the correct document identification and then if necessary, a second correlation to determine authentication such as by performing a 1 pixel shift multiple autocorrelation comparison of the two CDPs.

Abstract: A method for driving virtual facial expressions by automatically detecting facial expressions of a face image is applied to a digital image capturing device. The method includes the steps of detecting a face image captured by the image capturing device and images of a plurality of facial features with different facial expressions to obtain a key point position of each facial feature on the face image; mapping the key point positions to a virtual face as the key point positions of corresponding facial features on the virtual face; dynamically tracking the key point of each facial feature on the face image; estimating the key point positions of each facial feature of the current face image according to the key point positions of each facial feature on a previous face image; and correcting the key point positions of the corresponding facial features on the virtual face.

Abstract: In a method of evaluating an image for a dominant line, warped mesh information regarding saliency of the image is received. Displacements of components of the warped mesh information are determined relative to corresponding components in un-warped mesh information. Statistical representations of the displacements are utilized to evaluate the image for a dominant line.

Abstract: A system for screening cargo containers is provided including an image generation device, an apparatus for screening cargo containers and an output module. The image generation device generates an image signal conveying information related to the contents of the cargo container. The apparatus receives the image signal and a list of objects conveying objects expected to be present in the cargo container. A processing unit processes the image signal in combination with the list of objects and a group of target images associated with objects to derive mismatch information data. The mismatch information data conveys at least one distinction between the list of objects and the information related to the contents of the cargo container conveyed by the image signal. Information conveying the mismatch information data is released and conveyed to a user of the system by an output module.

Abstract: For each of a number of landmarks in an image an initial position of the landmark is defined. Next a neighborhood around the initial position, comprising a number of candidate locations of the landmark is sampled and a cost is associated with each of the candidate locations. A cost function expressing a weighted sum of overall gray level cost and overall shape cost for all candidate locations is optimized. A segmented anatomic entity is defined as a path through a selected combination of candidate locations for which combination the cost function is optimized.

Abstract: The present invention is directed to diagnostic imaging of small pulmonary nodules. There are two main stages in the evaluation of pulmonary nodules from Computed Tomography (CT) scans: detection, in which the locations of possible nodules are identified, and characterization, in which a nodule is represented by measured features that may be used to evaluate the probability that the nodule is cancer. Currently, the most useful prediction feature is growth rate, which requires the comparison of size estimates from two CT scans recorded at different times. The present invention includes methods for detection and feature extraction for size characterization. The invention focuses the analysis of small pulmonary nodules that are less than 1 centimeter in size, but is also suitable for larger nodules as well.

Abstract: Methods and apparatus, including systems and computer program products, for creating high dynamic range images. One method includes actions of establishing a current set of digital image; displaying a high dynamic range image and information identifying the current set of images, the high dynamic range image being computed from the current set of images; and repeatedly, as long as the user provides user input changing the current set of images, receiving user input changing the current set of images and in response to each user input displaying a high dynamic range image and information identifying the current set of images, where the high dynamic range image is computed from the current set of images.

Abstract: Disclosed is a method of simulating a dynamic object comprising a plurality of vertices. The method comprises defining a current position and a current velocity for a vertex “v” among the plurality of vertices, generating an estimated next position for vertex “v” based on the current position and current velocity, updating the estimated next position based on a plurality of constraints, and after updating the estimated next position, computing a next position and a next velocity for vertex “v” based on the current position and estimated next position.

Abstract: An image processing apparatus is provided that is capable of executing identification of an object by using a two-dimensional-image having a relatively low resolution. In the image processing apparatus, image data obtained by photo capturing a prescribed object is used as a processing target; with respect to at least one plane relating to a part of the object, an N-dimensional estimated feature value (N ?3) defining the plane is operated; the N-dimensional estimated feature value and information identifying the original object are associated with each other and stored as a recognition database in a storage unit; and the recognition database is applied to a recognition process of the object.

Abstract: A method determines a warping function between a first image and a second image. For each pixel in each scan-line of the first image, a distance matrix is constructed which represents distances between the pixel and a plurality of corresponding pixels of a second image. The distance matrices are ordered according a scan-line order of the pixels.

Abstract: A system and method for voice recognition is disclosed. The system enrolls speakers using an enrollment voice samples and identification information. An extraction module characterizes enrollment voice samples with high-dimensional feature vectors or speaker data points. A data structuring module organizes data points into a high-dimensional data structure, such as a kd-tree, in which similarity between data points dictates a distance, such as a Euclidean distance, a Minkowski distance, or a Manhattan distance. The system recognizes a speaker using an unidentified voice sample. A data querying module searches the data structure to generate a subset of approximate nearest neighbors based on an extracted high-dimensional feature vector. A data modeling module uses Parzen windows to estimate a probability density function representing how closely characteristics of the unidentified speaker match enrolled speakers, in real-time, without extensive training data or parametric assumptions about data distribution.

Abstract: A system includes a stereoscopic camera to generate images of a document, such as a book, that can be converted to a three-dimensional contour of the document. A second camera take a two-dimensional image of the document. A model is generated that defines a mapping between the three-dimensional contour of the document and a two-dimensional flattened (de-warped) version of the document. The model may be applied to the two-dimensional image taken by the second camera to compensate for warping in the image.

Abstract: A method and system for increasing the certainty of a silhouette matching process, where the process is being used for attitude determination of an object of interest, for example an aircraft. The method involves using one or more mask images that include structure or features that may or may not always be associated with the object of interest, and overlaying the mask image(s) onto a library image of the aircraft. Each pixel of the library image is compared against corresponding pixels of the mask image(s) to determine which pixels represent ambiguous areas of the library image. Those pixels are eliminated from consideration in determining a Fit score, where the Fit score represents a percentage value indicative of a certainty of the matching process in identifying the attitude of the aircraft. The method and system is applicable to a wide ranging variety of object detection applications.

Abstract: A method for creating a novel viewpoint image from a plurality of images includes the steps of simultaneously and continuously acquiring sets of images from a plurality of cameras. Predetermined background based correspondence fields are used to detect novel objects. Image representations are assigned for these objects likely new correspondences. These likely new correspondences are tested and further improved upon in a refinement step. The resulting correspondences are used to construct a novel viewpoint image.

Abstract: An image matching method includes setting first and second lattices to first and second images respectively, computing potential force to each second lattice point of the second lattice by a gradient of an image correlation potential energy based on a position of each first lattice point and pixel thereof and a position of the second lattice point and pixel thereof, computing elasticity force to the second lattice point from elasticity energy between the second and adjacent lattice points, computing frictional force occurring at the second lattice point, performing a numerical analysis of an equation of motion regarding the second lattice point and based on the potential force, elasticity force and frictional force to obtain a convergence state of the second lattice points, and adding a new lattice point between an adjacent lattice point pair of second lattice points according to a distance between the adjacent lattice point pair.

Abstract: A method of atlas mapping and an apparatus for image mapping are provided. The method includes generating a four dimensional template, linearly transforming the four dimensional representation of a subject to the four dimensional template, generating attribute vectors for a plurality of points of the linearly transformed four dimensional representation of the subject and of the four dimensional template, the plurality of points consisting of a plurality of temporally successive voxels of the four dimensional representation of the subject and of the four dimensional template, evaluating the generated attribute vectors to identify points in the four dimensional template that correspond to the four dimensional representation of the subject, and deforming the four dimensional template to the four dimensional representation of the subject according to relative positions of the corresponding points of the linearly transformed four dimensional representation of the subject and of the four dimensional template.

Abstract: Techniques and tools for video frame interpolation and motion analysis are described. The techniques and tools may be implemented separately or in combination in software and/or hardware devices for various applications. For example, a media playback device uses frame interpolation and motion analysis in real time to increase the frame rate of streamed video for playback. The device uses feature/region selection in global motion estimation, local motion estimation to correct the global motion estimation at an intermediate timestamp for a synthesized frame, and vector switching in the local motion estimation.

Abstract: A target recognizing device that creates a template concerning a target by itself is provided. The device comprises a robot camera for creating continuous images showing a target and an arithmetic operation unit for controlling the robot camera. A robot camera acquires continuous image information by means of a camera rotatable by a motor. The arithmetic operation unit comprises a template management section for allowing an update part to update a template stored in a template storage part and a target recognition section having an image storage part where the acquired image information is stored, a recognizing part for recognizing a target by using the template, and a motion command part for rotating a camera according to the result of recognition.

Abstract: A system and method to facilitate pattern recognition or matching between patterns are disclosed that is substantially invariant to small transformations. A substantially smooth deformation field is applied to a derivative of a first pattern and a resulting deformation component is added to the first pattern to derive a first deformed pattern. An indication of similarity between the first pattern and a second pattern may be determined by minimizing the distance between the first deformed pattern and the second pattern with respect to deformation coefficients associated with each deformed pattern. The foregoing minimization provides a system (e.g., linear) that may be solved with standard methods.

Abstract: A system and method for non-rigid image registration using distance functions includes portions for receiving a source shape into an image space, integrating a global linear registration model with local deformations to assess the source shape, optimizing a functional defined on a parameter or feature space where the functional quantifies the similarity between the source shape and a target shape in terms of distance functions, creating an augmented registration space including a plurality of target shape clones coherently positioned in the image space, tracking moving interfaces between the source shape and the target shapes and clones through a level set method, and registering the source shape by seeking mutual correspondences between the source shape, the target shape and the target shape clones.

Abstract: The present invention relates to an iris extraction method. In the method, two searching regions are defined in a face image. A deformable template match algorithm and an energy function are used to measure the energy of each pixel with different hypothetical circular templates within the searching region. Pixels with the same hypothetical radius having energies greater than a predetermined threshold are recorded as iris candidates, wherein the pixel having the maximal energy is recorded as first iris candidate. Further, it detects associated iris pairs from iris candidates in each searching region, records the lower iris candidate having the maximal energy of iris pairs as second iris candidate, and selects the best iris candidate from first iris candidate and second iris candidate. Finally, it designates the best iris candidate having the maximal energy of all best iris candidates with different hypothetical radius as the iris in the face image.

Abstract: The invention establishes a target face to which a texture will be mapped and determines the start point for drawing the texture and the number of drawing iterations. In accordance with the drawing start point and the number of drawing iterations, the texture is mapped to the target face and, further, a wireframe dividing the mapped texture is drawn. When movement of the wireframe is requested, the drawing start point is changed in accordance with the amount of movement of a drag operation. When the number of lines of the wireframe is requested, the number of drawing iterations is changed in accordance with the amount of movement of a drag operation.

Abstract: An image processor that calculates values that are related to distortion between two image parts. The values are detected in a previous calculation. Those values are then used in the next calculation cycle to detect an early exit. That value, called least, divided by the number of accumulators, and its negative is loaded into the accumulators. When the accumulators reach zero, an early exit is established.

Abstract: Computerized method for studying alterations in the location of an object, said method comprising the steps of computing successive shapes by means of a dynamic contour model algorithm, whereby the dynamic contour model algorithm develops successive shapes and determines a correlation between two successive shapes and for stopping the reiteration of the algorithm when said determined correlation is lower than a predetermined value.

Abstract: A system and method to facilitate pattern recognition or matching between patterns are disclosed that is substantially invariant to small transformations. A substantially smooth deformation field is applied to a derivative of a first pattern and a resulting deformation component is added to the first pattern to derive a first deformed pattern. An indication of similarity between the first pattern and a second pattern may be determined by minimizing the distance between the first deformed pattern and the second pattern with respect to deformation coefficients associated with each deformed pattern. The foregoing minimization provides a system (e.g., linear) that may be solved with standard methods.

Abstract: It is controlled by a non-transformation coefficient calculated by a non-transformation calculation section, whether feedback is performed or not. When the feedback is ended, a collation value calculation section calculates a collation value (coincidence coefficient×non-transformation coefficient). A display process section compares the collation value with a predetermined threshold value. When the collation value is a threshold value or more, a signature is decided as a true signature, and the transformation-corrected image and the reference image are displayed such that these images overlap.

Abstract: Control points used in deriving an object boundary for a prior frame are overlaid onto a current frame. An initial estimate of an object boundary are derived from the control points and edge energy data. The operator adjusts the control points to better model the boundary for the current frame. For each updated control point, the object boundary is rederived. A restricted area is defined encompassing the initial control points. When a control point is moved outside the restricted area, the restricted area is redefined to accommodate it. The boundary between control points is derived by finding a best path. Only points within the restricted area are considered. A first set of rules is used to find the best path when the distance between the two points is less than threshold value. A second set of rules is used when the distance between the two points exceeds the threshold value.

Abstract: A system and method for detecting a shape in an image are provided. The method comprises: constructing a deformable model from an image; deforming the deformable model to remove an undesired shape in a portion of the image; computing properties of the deformed model to enable detection of a desired shape in the portion of the image; and detecting the desired shape based on the computed properties.

Abstract: A system and method to facilitate pattern recognition or matching between patterns are disclosed that is substantially invariant to small transformations. A substantially smooth deformation field is applied to a derivative of a first pattern and a resulting deformation component is added to the first pattern to derive a first deformed pattern. An indication of similarity between the first pattern and a second pattern may be determined by minimizing the distance between the first deformed pattern and the second pattern with respect to deformation coefficients associated with each deformed pattern. The foregoing minimization provides a system (e.g., linear) that may be solved with standard methods.

Abstract: A system and method for utilizing a search tool that registers transformation of a trained pattern by at least four degrees of freedom to register the instance of a pattern in an arbitrary six-degree-of-freedom pose is provided. The search tool is first trained to recognize a plurality of versions of a trained pattern/fiducial that are incrementally transposed to induce differing levels of known aspect and shear. An object having several instances of the trained pattern located at known spacings and orientations therebetween is imaged by the trained search tool, and the located instances of the trained pattern are compared to expected instances of the trained pattern to measure relative six-degree-of-freedom orientation for the underlying object.

Abstract: A graphical user interface (GUI) is used to quickly and easily find data patterns within a data sequence that match a target data pattern representing an event of interest. The user first specifies the target data pattern either by dragging a cursor over the target or by selecting the target from a palette of predefined trend patterns. Search criteria, such as a match threshold and amplitude and duration constraints, are then specified. A pattern recognition technique is then used to find data patterns within the data sequence that satisfy the search criteria. These matching data patterns are then presented to the user ranked by similarity, avoiding the need for the user to sift through large amounts of irrelevant data. Particular pattern recognition algorithms are also disclosed.

Abstract: The present invention is directed to an imaging apparatus for examining an object having smooth surfaces to determine shape deformations in the surface of object. The imaging apparatus includes an imaging device for obtaining a scanned image of the object to be examined. A reference image of the object is stored in a memory. An image register is coupled to the imaging device and to the memory containing the reference image of the object. The image register stores patch information corresponding to both the reference image and the scanned image. A transformation estimator is coupled to the image register, and provides a transform for registering the scanned image to the reference image. A deformation estimator is coupled to the transformation estimator and to the image register. The deformation estimator is configured to utilize the transform and the patch information to determine shape deformations of the object.

Abstract: An accumulation method for fast pattern search can accurately locate regular shaped patterns of interest. The method can be used for invariant search to match patterns of interest in images where the searched pattern varies in size or orientation or aspect ratio, when pattern appearance is degraded, and even when the pattern is partially occluded, where the searched image is large, multidimensional, or very high resolution, or where the pattern size is large. The accumulation transformation of the input image is determined based upon the searched pattern. The difference between the accumulation image and a shifted version of itself forms a basic shape pattern image that indicates potential location of the pattern within the input image. The computation requirement is independent of the size of the pattern region.

Abstract: A digital image signal is watermarked by locally changing geometric features of the image. The watermark consists of a pseudo-random, dense subset of image pixels, e.g., a pattern of lines (20). A number of significant image pixels (21,22,23), i.e., pixels which give the highest response to a predetermined processing operation, is determined and then moved (24) to the vicinity (&dgr;) of the line pattern. As a result of this “warping”, the majority of significant image pixels (21,22) is eventually located within the vicinity of the line pattern. At the receiver end, the most significant pixels of an input image are again determined. The image is a watermarked image if a statistically high percentage lies within the vicinity of the line pattern.

Abstract: An image processor that calculates values that are related to distortion between two image parts. The values are detected in a previous calculation. Those values are then used in the next calculation cycle to detect an early exit. That value, called least, divided by the number of accumulators, and its negative is loaded into the accumulators. When the accumulators reach zero, an early exit is established.

Abstract: A system and method for non-rigid image registration using distance functions includes portions for receiving a source shape into an image space, integrating a global linear registration model with local deformations to assess the source shape, optimizing a functional defined on a parameter or feature space where the functional quantifies the similarity between the source shape and a target shape in terms of distance functions, creating an augmented registration space including a plurality of target shape clones coherently positioned in the image space, tracking moving interfaces between the source shape and the target shapes and clones through a level set method, and registering the source shape by seeking mutual correspondences between the source shape, the target shape and the target shape clones.

Abstract: The present invention provides an image recognition method whereby the influence of noise due to fluctuations of the ambient environment can be minimized and which has good response and high recognition accuracy, without using binary processing.