Abstract: Methods for locating a feature on geospatial imagery and systems for performing those methods are disclosed. An accuracy level of each of a plurality of geospatial vector datasets available in a database can be determined. Each of the plurality of geospatial vector datasets corresponds to the same spatial region as the geospatial imagery. The geospatial vector dataset having the highest accuracy level may be selected. When the selected geospatial vector dataset and the geospatial imagery are misaligned, the selected geospatial vector dataset is aligned to the geospatial imagery. The location of the feature on the geospatial imagery is then determined based on the selected geospatial vector dataset and outputted via a display device.

Abstract: Embodiments of the present invention characterizing the uncertainty of the orbital state of an Earth-orbiting space object hereof using a Gauss von Mises probability density function defined on the n+1 dimensional cylindrical manifold n×. Additionally, embodiments of the present invention can include transforming a Gauss von Mises distribution under a diffeomorphism and approximating the output as a Gauss von Mises distribution. Embodiments of the present invention can also include fusing a prior state represented by a Gauss von Mises distribution with an update report, wherein the update can be either another Gauss von Mises distribution of the same dimension as the prior or an observation related to the prior by a stochastic measurement model. A Gauss von Mises distribution can be calculated from a plurality of reports, wherein the reports are either Gauss von Mises distributions or observations related to the state space by a stochastic measurement model.

Abstract: A computer implemented method and an apparatus for estimating head position of an individual are described. The method includes cropping an image object in an image. The image object may include a visual representation of a head of at least one individual. The method further includes abstracting one or more features from the image object. Subsequently, position of the head of the at least one individual in the image object is determined using the one or more features and a predefined statistical model. The predefined statistical model may include a value of each of the pixels of one or more stored images representing variety of head poses and ranges of head poses of the at least one individual.

Abstract: In some embodiments, a system and method for estimating the geographical location at which image data was captured with a camera identifies matching feature points between the captured images, estimates a pose of the camera during the image capture from the feature points, performs geometric reconstruction of a scene in the images using the estimated pose of the camera to obtain a reconstructed scene, and compares the reconstructed scene to overhead images of known geographical origin to identify potential matches

Abstract: Methods, devices, and systems for determining feature descriptors are provided. An example includes defining a plurality of anchor points within a patch of pixels in a particular area that includes a detected feature in an image, defining a first set of subpatches and calculating an intensity of each of the first set of subpatches, defining a second set of subpatches and calculating an intensity of each of the second set of subpatches, comparing the intensity of each of the second set of subpatches to the intensity of each of the first set of subpatches and if the intensity of a second set subpatch is higher than the intensity of a first set subpatch assign a binary value, otherwise assign an alternative binary value, and concatenating all the assigned binary values into a binary feature descriptor.

Abstract: A target tracking apparatus and method according to an exemplary embodiment of the present invention may quickly and accurately perform target detection and tracking in a photographed image given as consecutive frames by acquiring at least one target candidate image most similar to a photographed image of a previous frame among prepared reference target images, determining one of the target candidate images as a target confirmation message based on the photographed image, and calculating a homography between the determined target confirmation image and the photographed image, and searching the photographed image of the previous image for feature points according to the calculated homography, and tracking an inter-frame change from the previous frame of the found feature points to a current frame.

Abstract: The present invention relates to an apparatus and method for efficiently generating feature data which properly determines a feature point indicating features of images and describes the feature point. The apparatus for generating image feature data comprises: a feature point determining unit which determines a feature point from an image and extracts information on the determined feature point; a feature point filtering unit which determines, as a final feature point, at least one feature point from among the feature points determined by the feature point determining unit; and a feature data generating unit which generates image feature data based on the final feature points determined by the feature point filtering unit and feature point information on the final feature points.

Abstract: A computer implemented method for matching video data to a database containing a plurality of video fingerprints of the type described above, comprising the steps of calculating at least one fingerprint representing at least one query frame from the video data; indexing into the database using the at least one calculated fingerprint to find a set of candidate fingerprints; applying a score to each of the candidate fingerprints; selecting a subset of candidate fingerprints as proposed frames by rank ordering the candidate fingerprints; and attempting to match at least one fingerprint of at least one proposed frame.

Abstract: In one aspect, a method to reduce false alarms in identifying whether a candidate image is from an object class includes projecting the candidate image onto an object class subspace and projecting the candidate image onto a non-object class subspace. The method also includes determining whether the candidate image is from the object class using a Bayesian decision function based on the projections on the object class subspace and the non-object class subspace. In another aspect, a method to reduce false alarms in identifying whether a candidate mine image is from a mine class includes projecting the candidate mine image onto a mine subspace and projecting the candidate mine image onto a non-mine subspace. The method also includes determining whether the candidate mine image represents a mine using a Bayesian decision function based on the projections on the mine class subspace and the non-mine class subspace.

Abstract: The present invention relates to a device and method for analyzing the correlation between an image and another image or between an image and a video. The device for analyzing the correlation between images and the method for using same include: a feature data generating unit for determining a feature point of an image and generating feature data which includes feature point orientation information on each determined feature point; and a relation analyzing unit for analyzing the correlation between an image and another image using feature data generated from the feature data generating unit.

Abstract: The invention relates to a method and a system for estimating the resemblance between two images of optionally different modalities. More particularly, the invention makes it possible to characterize a similarity between two binary images formula (I): it allows in particular registration of images acquired in the fields of teledetection, medical imaging, industrial vision.

Abstract: A method and system for measuring and determining the full-field spatial distributions of strain tensor field components in a two or three-dimensional space, as a consequence of deformation under generalized loading conditions. One or more digital cameras may be used to acquire successive images of a deforming body with optically distinctive features on its surface. A method for determining the location of characteristic points of the surface features and another one for tracking these points as deformation occurs. Elongations between neighboring points on the vicinity of a location of interest are computed. The elongation between points is calculated even though discontinuities may exist between them. Strain tensor fields are directly calculated as a tensor approximation from these elongations without determining or using the displacement vector distributions.

Abstract: A method for comparing videos includes detecting local features in frames of a video and determining the positions of detected local features in a first frame compared to their respective positions in a succeeding second frame. For at least some of the detected local features, a movement indicator is generated representing the position of a local feature in the first frame relative to the position of that local feature in the second frame, the movement indicator including an orientation, to obtain a plurality of movement indicators. Movement indicators are grouped by orientation ranges. The number of grouped movement indicators in an orientation range are summed to obtain a value for that orientation range. A time series of values for at least one given orientation range is obtained by determining the positions of detected local features in a first frame compared to their respective positions in a succeeding second frame for a plurality of pairs of frames.

Abstract: An image processing device includes: a first arithmetic operation unit that individually calculates an evaluation value based upon first factors, in correspondence to each of a plurality of images included in an image batch input thereto; an exclusion unit that executes exclusion processing based upon calculation results provided by the first arithmetic operation unit so as to exclude an image, the calculation results for which satisfy a predetermined condition, from the image batch; a second arithmetic operation unit that individually calculates an evaluation value based upon second factors, which include a factor different from the first factors, in correspondence to each image in the image batch having undergone the exclusion processing; and an evaluation processing unit that ranks individual images in the image batch having undergone the exclusion processing based upon calculation results provided by the second arithmetic operation unit.

Abstract: A method and mobile terminal for correcting a gaze of a user in an image includes setting eye outer points that define an eye region of the user in an original image, transforming the set eye outer points to a predetermined reference camera gaze direction, and transforming the eye region of the original image based on the transformed eye outer points.

Abstract: Systems, methods, and media for optical recognition are provided. In some embodiments, systems for optical recognition are provided, the systems comprising: at least one hardware processor that: identifies a plurality of fixation points in optically detected data; identifies features of the plurality of fixation points; and identifies one or more characteristics of an object represented in the optically detected data. In some embodiments, methods for optical recognition are provided, the methods comprising: identifying a plurality of fixation points in optically detected data using a hardware processor; identifying features of the plurality of fixation points using the hardware processor; and identifying one or more characteristics of an object represented in the optically detected data using the hardware processor.

Abstract: A method of detecting a predefined set of characteristic points of a face from an image of the face includes a step of making the shape and/or the texture of a hierarchy of statistical models of face parts converge over real data supplied by the image of the face.

Abstract: A system and method are disclosed for detecting road marking in a video using learned road marking templates. The system comprises a template learning module configured to learn the feature-based road marking templates from a set of training images. The template learning module is configured to rectify each training image, detect multiple regions of interest, and for each detected region of interest, detect multiple key points. The template learning module extracts feature vectors for the detected key points and builds the road marking templates from the feature vectors. The system also includes a road marking detection module for detecting road markings in a video at runtime using the learned road marking templates. During runtime, these templates are matched using a two-step process of first selecting promising feature matches and subsequently performing a structural matching to account for the shape of the road markings.

Abstract: The outlines of specific subjects becoming unnaturally deformed is prevented, while reducing the amount of calculations. A reference image setting section sets a reference image which is to become a reference from among a plurality of images. A specific subject detecting section detects a specific subject from within the reference image. A feature point extracting section extracts a plurality of feature points from within the reference image such that the average density of the feature points become higher in the vicinity of the outline of the specific subject. A corresponding point obtaining means corresponding point obtaining section obtains corresponding points within the other images that correspond to the extracted feature points. A coordinate converting section converts the coordinates of the positions of each pixel within the reference image and/or the other images such that the positions of the feature points and the positions of the corresponding points match.

Abstract: A perimeter around a detected object in a frame of image data can be generated in a first coordinate system. The perimeter can be converted from the first coordinate system into a second coordinate system having the same aspect ratio as the first coordinate system. A first metadata entry can include dimensions of image data in the second coordinate system. A second metadata entry can provide a location and dimensions of the converted perimeter in the second coordinate space. Additional metadata can indicate matching objects between frames, position of an object relative to other objects in a frame, a probability that an object is correctly detected, and a total number of objects detected across multiple frames of image data.

Abstract: A computing device for motion detection in a system capable of detecting feature points of an object of interest is disclosed. The computing device includes a vector forming unit to form a plurality of vectors associated with a set of the feature points and form a vector set based on the vectors, a posture identifying unit to identify a match of a posture in a database based on the vector set, a motion similarity unit to identify a set of predetermined postures in the database based on the matched posture and an immediately previous matched posture, and a motion identifying unit to identify a predetermined motion in the database based on the set of predetermined postures.

Abstract: A feature point positioning apparatus positions a plurality of feature points for a predetermined pattern in image data. The apparatus selectively executes first candidate decision processing which decides position candidates of the feature points and second candidate decision processing which decides position candidates of the feature points at a higher processing speed than the first candidate decision processing in accordance with an operation mode, and executes, when the operation mode is a high-speed mode in which an operation is executed at a higher speed than a normal mode, the second candidate decision processing for more feature points than in the normal mode; and corrects the position candidates of the plurality of feature points obtained by the first candidate decision processing and the second candidate decision processing based on a layout relationship among the plurality of feature points.

Abstract: A variety of methods, systems, devices and arrangements are implemented for use with motion capture. One such method is implemented for identifying salient points from three-dimensional image data. The method involves the execution of instructions on a computer system to generate a three-dimensional surface mesh from the three-dimensional image data. Lengths of possible paths from a plurality of points on the three-dimensional surface mesh to a common reference point are categorized. The categorized lengths of possible paths are used to identify a subset of the plurality of points as salient points.

Abstract: A plurality of points with identical geometric feature is compared with their SEM characteristic features to inspect defect in a localized image. Original design information is included in the geometric feature such that absolute compare can be performed in this inspection method. Further, this method can also be applied to the localized image with or without repeated or redundant pattern.

Abstract: An information processing apparatus includes an image input unit which inputs image data containing a face, a face position detection unit which detects, from the image data, the position of a specific part of the face, and a facial expression recognition unit which detects a feature point of the face from the image data on the basis of the detected position of the specific part and determines facial expression of the face on the basis of the detected feature point. The feature point is detected at a detection accuracy higher than detection of the position of the specific part. Detection of the position of the specific part is robust to a variation in the detection target.

Abstract: A computer-implemented method for measuring full field deformation characteristics of a deformable body. The method includes determining optical setup design parameters for measuring displacement and strain fields, and generating and applying a dot pattern on a planar side of a deformable body. A sequence of images of the dot pattern is acquired before and after deformation of the body. Irregular objects are eliminated from the images based on dot light intensity threshold and the object area or another geometrical cutoff criterion. The characteristic points of the dots are determined, and the characteristic points are matched between two or more of the sequential images. The displacement vector of the characteristic points is found, and mesh free or other techniques are used to estimate the full field displacement based on the displacement vector of the characteristic points. Strain tensor or other displacement-derived quantities can also be estimated using mesh-free or other analysis techniques.

Abstract: A feature point matching device includes a position feature generation unit for generating a position feature quantity for a feature point based on a position of the feature point included in an image, a pixel value feature generation unit for generating a pixel value feature quantity for the feature point based on a pixel value of the feature point included in the image, and a feature matching unit for matching a first feature point and a second feature point based on a degree of matching between a first position feature quantity generated from a first image and a second position feature quantity generated from a second image and a degree of matching between a first pixel value feature quantity generated from the first image and a second pixel value feature quantity generated from the second image.

Abstract: Techniques are disclosed that involve the detection of smiles from images. Such techniques may employ local-binary pattern (LBP) features and/or multi-layer perceptrons (MLP) based classifiers. Such techniques can be extensively used on various devices, including (but not limited to) camera phones, digital cameras, gaming devices, personal computing platforms, and other embedded camera devices.

Abstract: A method for comparing a first image with a second image. The method identifies first keypoints in the first image and second keypoints in the second image and associates each first keypoint with a corresponding second keypoint to form a corresponding keypoint match. For each pair of first keypoints, the method further calculates the distance therebetween for obtaining a corresponding first length. Similarly, for each pair of second keypoints, the method calculates the distance therebetween for obtaining a corresponding second length. The method further calculates a plurality of distance ratios; each distance ratio is based on a length ratio between a selected one between a first length and a second length and a corresponding selected one between a second length and a first length, respectively.

Abstract: A method for detecting features in digital numeric data comprises obtaining digital numeric data comprising values corresponding to a plurality of sampling points over a domain space having at least one dimension, computing a plurality of scale-space data comprising filtering said digital numeric data using a filter bank, determining a plurality of feature regions each corresponding to a local extremum in scale and location of the scale-space data; and determining a feature region descriptor for each of said plurality of feature regions. The filter bank is a Cosine Modulated Gaussian filter bank in which the standard deviation parameter of the Gaussian equals 1 ? ? ln ? ( 2 ) 2 ? 2 b + 1 2 b - 1 multiplied by the cosine wavelength, in which b is in the range of 0.75 to 1.25, or said filter bank is an Nth-order Gaussian Derivative filter bank with N being in the range of 5 to 20.

Abstract: A method of finding point correspondences includes dividing a first left image into first sub regions and a first right image into second sub regions, detecting first feature points according to first threshold values respectively corresponding to the first sub regions and detecting second feature points according to the first threshold values respectively corresponding to the second sub regions, matching each of the first feature points with each of the second feature points, finding point correspondences in the matched feature points, and providing the point correspondences to an image processing device.

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: A hardware coprocessor architecture calculates the Difference-of-Gaussian (DoG) pyramid of an input image and extracts from this the interest points to be used in several image detection algorithms. Advantages of the architecture include the possibility to process the image by stripes, namely by blocks having one dimension coincident with the input image width, in the absence of an input frame buffer and the possibility to avoid RAM memory. The coprocessor is suitable to be tightly coupled with raw image sources like sensors.

Abstract: A method for guiding a user with a suitable composition includes detecting scene information from an input image, recognizing a scene of the input image by using the detected scene information, extracting composition information corresponding to the recognized scene, and displaying the extracted composition information. Accordingly, an apparatus for guiding a user with a suitable composition and a digital photographing apparatus using the method displays a composition suitable for a current scene by automatically recognizing the current scene, so that the user photographs the current scene quickly and easily.

Abstract: Various methods for feature extraction using local primitive code are provided. One example method includes determining an origin block within an image, an arc, an arc orientation, and a distance between the origin block and the arc. In this regard, the origin block may include one or more pixels of the image. The example method also includes determining at least one arc block. In this regard, the arc block may be located on the arc and the arc may be located with respect to the origin block based on the arc orientation and the distance. Some or all of the at least one arc blocks may include one or more pixels of the image. The example method also includes determining at least one feature value based on attributes of the origin block and the at least one arc block. Similar and related example methods and example apparatuses are also provided.

Abstract: Image-processing apparatus and methods to adaptively vary an interest point threshold value and control a number of interest points identified in an image frame are described. Sub-regions of an image frame may be processed in a sequence, and an interest point threshold value calculated for each sub-region. The calculated value of the interest point threshold may depend upon pre-selected values and values determined from the processing of one or more prior sub-regions. By using adaptive thresholding, a number of interest points detected for each frame in a sequence of image frames may remain substantially constant, even though objects within the frames may vary appreciably.

Abstract: An image processing method includes sending test image data to a plurality of image recognition units configured to detect a recognition object from an image, setting an evaluation condition for evaluating a recognition result, evaluating a recognition result of the test image data by each of the plurality of image recognition units under the evaluation condition, and selecting from the plurality of image recognition units an image recognition unit to be used based on an evaluation result by the evaluation.

Abstract: A method, apparatus, and system are described for model-based playfield registration. An input video image is processed. The processing of the video image includes extracting key points relating to the video image. Further, whether enough key points relating to the video image were extracted is determined, and a direct estimation of the video image is performed if enough key points have been extracted and then, a homograph matrix of a final video image based on the direct estimation is generated.

Abstract: An image processing method and a system for processing the same are provided. The image processing method includes the following steps. A first image having several first areas and a second image having several second areas are provided. Each first area has a first feature point having the largest or the smallest grey value in the first area. Each second area has a second feature point having the largest or the smallest grey value in the second area. A first relationship between the first feature points and a second relationship between the second feature points are created. The first and the second feature points are paired by a microprocessor according to the first and the second relationship.

Abstract: When generating outline data, the contour pixels of the binarized image data are first extracted. Based on the extracted contour pixels, the contour of the image data is approximated to a straight line. In the straight-line approximation process, the distance between adjacent contour pixels among the extracted contour pixels is calculated. It is determined based on the result of the comparison between the length of the first straight line and the length of the second straight line whether the first straight line is used as a contour of the image data. The first straight line connects the first contour pixel and the second contour pixel adjacent to the first contour pixel. The second straight line connects the second contour pixel and the third contour pixel adjacent to the second contour pixel.

Abstract: The present invention relates to an apparatus and method for detecting and recognizing an object in an image, the method comprising a plurality of stages, wherein at least one of the stages comprises an integrated approach of feature detection and object recognition of at least a part of the object. In a further embodiment at least one of the stages comprises identifying an image part that contains a feature point, and matching the image part to a set of hierarchies of templates, wherein a hierarchy comprises templates for an object to recognized, a template describes at least a part of an object to be recognized, and a child template describes a sub-part of the part of the object described by its parent template.

Abstract: A method for generating a projection image from three-dimensional image data, including the steps of determining first three-dimensional image data to be projected from the three-dimensional image data, generating a first projection image by projecting the first three-dimensional image data in a first projection direction by maximum intensity projection or minimum intensity projection, specifying a first coordinate position in the first projection image, determining second three-dimensional image data to be projected from the three-dimensional image data, generating a second projection image by projecting the second three-dimensional image data in a second projection direction different from the first projection direction, and displaying a second coordinate position, in the second projection image, obtained by projecting a first voxel in the three-dimensional image data, which corresponds to the pixel at the first coordinate position in the first projection image, onto the second projection image from the sec

Abstract: A method for identifying digital images having matching backgrounds from a collection of digital images, comprising using a processor to perform the steps of: determining a set of one or more feature values for each digital image in the collection of digital images, wherein the set of feature values includes an edge compactness feature value that is an indication of the number of objects in the digital image that are useful for scene matching; determining a subset of the collection of digital images that are good candidates for scene matching by applying a classifier responsive to the determined feature values; applying a scene matching algorithm to the subset of the collection of digital images to identify groups of digital images having matching backgrounds; and storing an indication of the identified groups of digital images having matching backgrounds in a processor-accessible memory.

Abstract: Provided are a method and apparatus for simultaneously reducing various kinds of noises in an image output from an image sensor equipped to a camera. The image noise reducing method includes: determining a color reference value, which corresponds to a corrected value of the color value of a current pixel, based on the difference values between the color value of the current pixel among pixels of an image and the color values of neighboring pixels around the current pixel; determining unified filter kernel coefficients, which represent the weights of the current pixel and the neighboring pixels, based on the similarities between the determined color reference value and the color values of the neighboring pixels; and reducing a noise in the color value of the current pixel by weighted-averaging the color value of the current pixel and the color values of the neighboring pixels using the unified filter kernel coefficients.

Abstract: There is provided a method of measuring 3D depth of a stereoscopic image, comprising providing left and right eye input images, applying an edge extraction filter to each of the left and right eye input images, and determining 3D depth of the stereoscopic image using the edge extracted left and right eye images. There is also provided an apparatus for carrying out the method of measuring 3D depth of a stereoscopic image.

Abstract: An image processing apparatus for searching for a feature point by use of a depth image and a method thereof are provided. The image processing apparatus includes an input unit configured to input a three-dimensional image having depth information, a feature point extraction unit configured to obtain a designated point from an object image extracted from the depth image to obtain a feature point that is located at a substantially farthest distance from the designated point, and to obtain other feature points that are located at substantially farthest distances from feature points that are previously obtained as well as the designated point. The apparatus includes a control unit configured to control the input unit and the feature point extraction unit so that time in estimating a structure of the object is reduced, and a recognition result is enhanced.

Abstract: A system, apparatus and method of obtaining data from a 2D image in order to determine the 3D shape of objects appearing in said 2D image, said 2D image having distinguishable epipolar lines, said method comprising: (a) providing a predefined set of types of features, giving rise to feature types, each feature type being distinguishable according to a unique bi-dimensional formation; (b) providing a coded light pattern comprising multiple appearances of said feature types; (c) projecting said coded light pattern on said objects such that the distance between epipolar lines associated with substantially identical features is less than the distance between corresponding locations of two neighboring features; (d) capturing a 2D image of said objects having said projected coded light pattern projected thereupon, said 2D image comprising reflected said feature types; and (e) extracting: (i) said reflected feature types according to the unique bi-dimensional formations; and (ii) locations of said reflected feature

Abstract: Apparatuses, methods, systems and computer-readable media for generating a visibility map in determination of a user's location are presented. Some embodiments may identify a point of interest (POI) whose location is predefined on a map. The map may be an overhead map, similar to a map found at a shopping mall or a 3-dimensional map. Embodiments may then determine a plurality of vectors emanating from the POI. The vectors may be representative of lines of sight from the POI on the map. Embodiments may then integrate at least some of the plurality of vectors to calculate an area on the map representative of a visibility map.

Abstract: Three-dimensional coordinates of feature points of an object to be measured are back-projected to a frame image photographed from a specific position, and image coordinates of the back-projected feature points and the feature points in this frame image are compared. In this case, the feature points, which are mismatched, are removed as feature points which are mistracked between plural frames. In this case, two processing systems, of which initial conditions of calculation for obtaining the back-projected coordinates are different from each other, are performed, and the detection of the above mistracked points is performed on each of the two back-projected coordinates. The mistracked points detected in at least one of the processing systems are removed, and are not succeeded to the following processing.

Abstract: This disclosure describes techniques for searching for similar images to an image query by using an approximate k-Nearest Neighbor (k-NN) graph. The approximate k-NN graph is constructed from data points partitioned into subsets to further identify nearest-neighboring data points for each data point. The data points may connect with the nearest-neighboring data points in a subset to form an approximate neighborhood subgraph. These subgraphs from all the subsets are combined together to form a base approximate k-NN graph. Then by performing more random hierarchical partition, more base approximate k-NN graphs are formed, and further combined together to create an approximate k-NN graph. The approximate k-NN graph expands into other neighborhoods and identifies the best k-NN data points. The approximate k-NN graph retrieves the best NN data points, based at least in part on the retrieved best k-NN data points representing images being similar in appearance to the image query.