Abstract: An image recognition algorithm includes a keypoints-based comparison and a region-based color comparison. A method of identifying a target image using the algorithm includes: receiving an input at a processing device, the input including data related to the target image; performing a retrieving step including retrieving an image from an image database, and, until the image is either accepted or rejected, designating the image as a candidate image; performing an image recognition step including using the processing device to perform an image recognition algorithm on the target and candidate images in order to obtain an image recognition algorithm output; and performing a comparison step including: if the image recognition algorithm output is within a pre-selected range, accepting the candidate image as the target image; and if the image recognition algorithm output is not within the pre-selected range, rejecting the candidate image and repeating the retrieving, image recognition, and comparison steps.

Abstract: The present invention is a system and a method for processing stereo images utilizing a real time, robust, and accurate stereo matching system and method based on a coarse-to-fine architecture. At each image pyramid level, non-centered windows for matching and adaptive upsampling of coarse-level disparities are performed to generate estimated disparity maps using the ACTF approach. In order to minimize propagation of disparity errors from coarser to finer levels, the present invention performs an iterative optimization, at each level, that minimizes a cost function to generate smooth disparity maps with crisp occlusion boundaries.

Abstract: A system and method for determining if a computer user is a human or a computer program. A computer user is presented with a series of images, the series of images including: at least two related images of an object, at least one of the at least two related images independently processed using an image modifying algorithm; and at least one unrelated image. The computer user is then requested to identify the related images from the series of images. These image identifications are received from the computer user. A determination is made that the computer user is: a human if the image identifications only identify the related images from the series of images; and a non-human if the image identifications do not only identify the related images from the series of images.

Abstract: An imaging device and system include integration of an imaging camera (visible-light or near-infrared) with a thermal infrared sensor, capturing a baseline image from this camera simultaneously with acquisition of baseline thermal infrared data, which may correspond to a known good condition or part, automatic generation of an edge version of the baseline image for use as an alignment template, and on subsequent thermal infrared inspections, superposition of the alignment template on a live video image from the visible camera, so as to facilitate highly repeatable alignment of the thermal infrared sensor to an object being inspected.

Abstract: An apparatus and method for auto white balance adjusting using an effective area. The auto white balance adjusting apparatus may extract a grey area from an input image, and select an effective area from the grey area according to color temperature and luminance, thereby performing white balance using the effective area. The auto white balance adjusting apparatus may perform white balance of the input image using image information of the effective area which is a more precise grey area, thereby improving the white balance.

Abstract: A method and system for detection of video segments in compressed digital video streams is presented. The compressed digital video stream is examine to determine synchronization points, and the compressed video signal is analyzed following detection of the synchronization points to create video fingerprints that are subsequently compared against a library of stored fingerprints.

Abstract: Provided is a three-dimensional model classification method of classifying constitutions.

Abstract: In one embodiment, a system and method for modeling a three-dimensional object includes capturing two-dimensional images of the object from multiple different viewpoints to obtain multiple views of the object, estimating slices of the object that lie in parallel planes that cut through the object, and computing a surface of the object from the estimated slices.

Abstract: A method of analyzing a captured image comprising an instance of a target object comprises the steps of: for each of a plurality of different brightness threshold levels, generating contours from the captured digital image that indicate where in the captured digital image the pixel values of the captured digital image cross the respective brightness threshold level; identifying instances of a contour corresponding to a characteristic feature of said target object, the instances being detected at substantially similar image positions in the contours derived using at least two of the respective brightness threshold levels; and estimating a homography which maps the characteristic feature of the target object to its representation in the captured image, based upon the two or more instances of that target object's corresponding contour.

Abstract: A method for determining an absolute orientation in a global coordinate system is provided. In each grid position an image capturing device is directed to a selected object located in a scene. The method comprises: (A) reconstructing a scale model of the scene using point, line and/or region correspondences in multiple views of the scene obtained by the image capturing device placed in at least three grid positions; (B) obtaining location coordinates of at least three grid positions in the global coordinate system; and (C) determining a similarity transform from a photogrammetric coordinate system to the global coordinate system by determining rotation, translation and scale of the image capturing device for each grid position.

Abstract: An initial value is assigned to a center point for each cluster in a plurality of clusters. Each point in a point space is assigned to a closest cluster based on the distance between the each point and the center of nearest cluster. A first-assignment value is determined for each center point using the clusters the points are assigned to. A first-assignment dynamic validity index of a current cluster configuration is evaluated. Each point in the point space is reassigned to the closest cluster based on the first-assignment value of each center. A second-assignment value is determined for the center of each cluster according to the reassigning. A second-assignment dynamic validity index is evaluated using the second-assignment values. The current cluster configuration is selected if the difference between the dynamic validity indices is less than a threshold.

Abstract: The invention discloses a method and a system for generating a boundary in the process of rasterizing a vector graphic as well as a method for producing the system and a method for clipping the vector graphic. The method for generating a boundary includes: segment-polylining the vector graphic and obtaining data of the line segments; obtaining data of the end pixels according to the data of the line segments; obtaining data of boundary pixels according to the data of the end pixels and the relationship between the data of the end pixels; obtaining bitmap data of the boundary pixels according to the data of the boundary pixels. The algorithm used in the invention is very simple and efficient; at the same time, the generation of the bitmap data of the boundary pixels can be carried out synchronously with the generation of bitmap data of the graphic itself.

Abstract: A system detects an object contour with an image acquisition assembly, the object moving relative to the assembly. A line detector scans the surface line by line during a scan cycle, the line being transverse to the relative motion direction. During active periods, a light source emits light synchronized with the scan cycle, allowing the line detector to acquire a first group of at least one lit scan line. A second group of unlit scan line(s) is acquired during non-emitting idle periods. The object passes between the line detector and the light source. A processor receives and analyzes acquired scan lines. For each lit scan line group and a successive second unlit scan line group, the processor identifies a token pattern with a lit segment adjoining an unlit segment. The processor searches the first and second groups for the token pattern ending or reappearing to produce an object contour.

Abstract: A bearing is provided for use in connection with expansion joint systems. The bearing may be incorporated into expansion joint systems that are used in roadway constructions, bridge constructions, and architectural structures. The bearing can absorb increased loads that are applied to the expansion joint system. The structure of the bearing also permits improved motion of, and provides improved support for, the components of the expansion joint system that are supported on or engaged with the bearing.

Abstract: A system categorizes one or more objects based at least in part upon one or more characteristics associated therewith. A first classifier includes a rule set to determine if each of the one or more objects meets or exceeds a quality threshold. A second classifier, orthogonal to the first classifier, includes a rule set to determine if each of the one or more objects meets or exceeds a quality threshold. In one embodiment, the quality threshold associated with the first classifier and the quality threshold associated with the second classifier are less than a predetermined target threshold. The result for each object of the first classifier is compared to the result of the second classifier. The object is categorized if the result of the first classifier and the result of the second classifier match. The object is uncategorized if the result of the first classifier does not match the result of the second classifier.

Abstract: Quantized energy values are accessed to initially represent a temporally related group of content elements in a media sequence. The values are accessed over a matrix of regions into which the initial representation is partitioned. The initial representation may be downsampled and/or cropped from the content. A basis vector set is estimated in a dimensional space from the values. The initial representation is transformed into a subsequent representation, which is in another dimensional space. The subsequent representation projects the initial representation, based on the basis vectors. The subsequent representation reliably corresponds to the media content portion over a change in a geometric orientation thereof. Repeated for other media content portions of the group, subsequent representations of the first and other portions are averaged or transformed over time. The averaged/transformed values reliably correspond to the content portion over speed changes.

Abstract: An image taking apparatus according to an aspect of the invention comprises: an image pickup device which picks up an object image and outputs the picked-up image data; a face detection device which detects human faces in the image data; a face-distance calculating device which calculates the distance between the faces among a plurality of faces detected by the face detection device; and a controlling device which controls the image pickup device to start shooting, after a shooting instruction is issued, in the case where the distance between the faces calculated by the face-distance calculating device is not greater than a first predetermined threshold value. The image taking apparatus allows shooting the moment the distance between the faces is close enough not be greater than to a predetermined threshold value.

Abstract: A process for extracting iris data for biometric identification includes a thresholding method where the thresholds are selected according to a nonparametric approach that considers the grey scale and does not require classifying pixels as edge or non-edge pixels. An eye image is first acquired, where the eye image has component images including an iris image with an inner boundary and an outer boundary. The eye image has a distribution of grey levels. Component images, such as an iris image or a pupil image, from the eye image are segmented according to the distribution of grey levels. The inner boundary and outer boundary of the iris image are determined from the component images. The iris image within the inner boundary and outer boundary is processed for biometric identification. The component images may be segmented by creating an eye histogram of pixel intensities from the distribution of grey levels.

Abstract: Provided are methods for determining optimal features for classifying patterns or objects. Also provided are methods for image analysis. Further provided are methods for image searching.

Abstract: The different illustrative embodiments provide a method for identifying landmarks in an image. An image of a worksite is received. The image is analyzed to determine a suggested identity of a worksite feature in the image. The suggested identity of the worksite feature is sent over a communications unit. A confirmation of the suggested identity of the worksite feature is received to form a confirmed identity. The confirmed identity and a number of attributes associated with the confirmed identity is stored in a database.