Abstract: Methods, devices, and systems for use in accomplishing registration of a patient to a robot to facilitate image guided surgical procedures, such as stereotactic procedures.

Abstract: An image-based localization feature point registration apparatus includes a camera to capture an image, a feature point extractor to extract a feature point from the captured image, a calculator to calculate depth information about the feature point according to whether the feature point is one of two-dimensional (2D) and a three-dimensional (3D) corner, and a feature point register to register 3D coordinates of the feature point based on the depth information about the feature point and image coordinates of the feature point.

Abstract: Calibrated infrared and range imaging sensors are used to produce a true-metric three-dimensional (3D) surface model of any body region within the fields of view of both sensors. Curvilinear surface features in both modalities are caused by internal and external anatomical elements. They are extracted to form 3D Feature Maps that are projected onto the skin surface. Skeletonized Feature Maps define subpixel intersections that serve as anatomical landmarks to aggregate multiple images for models of larger regions of the body, and to transform images into precise standard poses. Features are classified by origin, location, and characteristics to produce annotations that are recorded with the images and feature maps in reference image libraries. The system provides an enabling technology for searchable medical image libraries.

Abstract: A computer implemented method for adaptive optical character recognition on a document with distorted characters includes performing a distortion-correction transformation on a segmented character of the document assuming the segmented character to be a candidate character. The method further includes comparing the transformed segmented character to the candidate character by calculating a comparison score. If the calculated score is within a predetermined range, the segmented character is identified with the candidate character. The method may be implemented in either of computer hardware configured to perform the method, or in computer software embodied in a non-transitory, tangible, computer-readable storage medium. Also disclosed are corresponding computer program product and data processing system.

Abstract: Apparatus for matching a query image against a catalog of images, comprises: a feature extraction unit operative for extracting principle features from said query image; a relationship unit operative for establishing relationships between a given principle feature and other features in the image, and adding said relationships as relationship information alongside said principle features; and a first comparison unit operative for comparing principle features and associated relationship information of said query image with principle features and associated relationship information of images of said catalog to find candidate matches.

Abstract: Systems and methods for estimating a posture of a body part of a user are disclosed. In one disclosed embodiment, an image is received from a sensor, where the image includes at least a portion of an image of the user including the body part. The skeleton information of the user is estimated from the image, a region of the image corresponding to the body part is identified at least partially based on the skeleton information, and a shape descriptor is extracted for the region and the shape descriptor is classified based on training data to estimate the posture of the body part.

Abstract: The present disclosure discloses simplified text classification with improved classification accuracy. The content of receive text is segmented to obtain multiple words. A sum of word vectors of the multiple words at a first level vector space of a vector space group is determined. A vector space in the vector space group is composed of one or more direct child classes that are non-leaf child classes in a tree class structure. The text is classified into a class among various class of the first level vector space that has a shortest distance to the sum of the word vectors.

Abstract: The present invention discloses a method for mapping and controlling the three-dimensional shrinkage and warpage occurring in rotomoulding applications.

Abstract: The invention discloses a detecting device for specific subjects and a learning device and method thereof. The detecting device for specific subjects includes an input unit, one or more strong classifying units, a storage unit and a judging unit, wherein the input unit is used for inputting images to be detected; the strong classifying units are used for carrying out strong classification to the image, each strong classifying unit includes one or more weak classifying units, and the weak classifying unit carries out weak classification to the image with a weak classifying template; the storage unit stores the weak classifying template used by the weak classifying unit; and the judging unit judges whether or not the image contains specific subjects according to the classification result of the strong classifying unit.

Abstract: An apparatus for providing an estimate for a 3D camera pose relative to a scene from 2D image data of 2D image frame provided by said camera. A candidate 2D key points detector determines candidate 2D key points from the 2D image frame. A detected 3D observations detector determines detected 3D observations from pre-recorded scene data and the candidate 2D key points. A detected 3D camera pose estimator determines a detected 3D camera pose estimate from the camera data, the detected 3D observations and the candidate 2D key points. A first storage stores the detected 2D candidate key points and the 2D image data, and outputs in response to a 3D camera pose estimate output previous 2D image data and candidate 2D key points related to a previous 3D camera pose estimate output. A second storage stores and outputs a previous 3D camera pose estimate.

Abstract: A device and apparatus for processing a depth-map including obtaining a depth-map based on a lossy compressed depth-map, the depth-map including depth information of a scene from a viewpoint, the scene includes an object; obtaining occlusion information for the scene from the viewpoint, the occlusion information including information occluded by the object in the depth-map; and processing at least part of the depth information using at least part of the occlusion information in order to reduce compression artifacts in the depth-map.

Abstract: Methods for post-processing of non-key frames of an image are described. According to the methods, reconstructed non-key frames are updated with information of neighboring key frames. Methods to evaluate whether to update or not update the non-key frames are also described.

Abstract: Disclosed is a method for extracting a region-of-interest of a user in a multimedia mobile terminal. The method includes: setting a search region in an input picture input to the multimedia mobile terminal on a screen picture-to-screen picture basis, and extracting a group of boundary of a region-of-interest based on a brightness difference between the set search region and an environment around the set search region; determining boundary coordinates of the region-of-interest through learning during a predetermined frame among the extracted group of boundary coordinate candidates; and enlarging the region-of-interest having the determined boundary coordinates, and displaying the enlarged region-of-interest. As a result, a region the user is most interested in or pays more attention to than other regions on a screen is automatically extracted.

Abstract: Two-dimensional image information and three-dimensional image information of a subject are acquired, facial recognition is performed using the two-dimensional image information to determine whether a recognized face is a registered user's face, an elliptical model of the user is matched to the three-dimensional image information to calculate an error if it is determined that the recognized face is the user's face, and it is determined whether the user's face is improperly used based on the error. The subject's face is determined using the two-dimensional image information and the three-dimensional image information of the subject and it is determined whether the recognized face is improperly used, thereby improving facial recognition reliability. Thus, information security is improved.

Abstract: A method for performing image recognition is disclosed. The method includes obtaining a collection of pixels and grouping at least some of the pixels into a set of cluster features based on gradient magnitude. For each cluster feature in the set, statistical variables are generated. The statistical variables represent a collective property of the pixels in the cluster feature. The statistical variables are utilized as a basis for comparing the collection of pixels to a different collection of pixels.

Abstract: In a pose estimation for estimating the pose of an object of pose estimation with respect to a reference surface that serves as a reference for estimating a pose, a data processing device: extracts pose parameters from a binarized image; identifies a combination of pose parameters for which the number of cross surfaces of parameter surfaces that accord with surface parameter formulas, which are numerical formulas for expressing a reference surface, is a maximum; finds a slope weighting for each of cross pixels, which are pixels on each candidate surface and which are pixels within a prescribed range, that is identified based on the angles of the tangent plane at the cross pixel and based on planes formed by each of the axes of parameter space; and identifies the significant candidate surface for which a number, which is the sum of slope weightings, is a maximum, as the actual surface that is the reference surface that actually exists in the image.

Abstract: An image processing apparatus may include a feature quantity extraction unit configured to extract a feature quantity from an image, a setting unit configured to set setting information including a plurality of setting items used to perform processing upon the image so that a designated setting item can be changed among the setting items; and a recording unit configured to associate the setting information with a feature quantity of the image and record them.

Abstract: Method for organizing a set of images into subsets of images, the method including the following procedures producing a respective model according to a plurality of feature points and the geometric relations between the feature points for each of the images, determining a similarity index between each pair of the images, according to the respective model of each image of the pair of the images, producing a distance matrix, according to the similarity index between each pair of the images, producing a set of coordinates, according to the distance matrix and sorting the images in plurality of dimensions, according to the set of coordinates.

Abstract: A method comprises segmenting a foreground and a background of an image; and extracting one or more features from the foreground and the background to recognize a brand image. The features comprises one or more from a group comprising a foreground area, coordinates of a foreground centeroid, a foreground symmetry, a connected property, a spatial moment of the foreground, a normalized center moment of the foreground, a background area, variations of the background in red, green and blue color channels, a ratio of the foreground area and the background, an entropy of the image, an edge density of the image.

Abstract: A multimodal biometric identification system captures and processes images of both the iris and the retina for biometric identification. Another multimodal ocular system captures and processes images of the iris and/or the from both eyes of a subject. Biometrics based on data provided by these systems are more accurate and robust than using biometrics that include data from only the iris or only the retina from a single eye. An exemplary embodiment emits photons to the iris and the retina of both eyes, an iris image sensor that captures an image of the iris when the iris reflects the emitted light, a retina image sensor that captures an image of the retina when the retina reflects the emitted light, and a controller that controls the iris and the retina illumination sources, where the captured image of the iris and the captured image of the retina contain biometric data.