Abstract: Systems and methods for character recognition by performing lateral view-based analysis on the character data and generating a feature vector based on the lateral view-based analysis.

Abstract: A metadata extraction machine accesses an image that depicts an item and identifies an attribute of the item and a descriptor that specifies a value for the identified attribute. The machine identifies the attribute and the descriptor based on an analysis of the accessed image that includes distinguishing a foreground portion of the image from a background portion of the image. Distinguishing a foreground portion from the background portion may include segmenting the image to identify the foreground and background portions. Some or all of the analysis may be performed by the machine. The machine then stores the identified descriptor of the item as metadata of the item.

Abstract: A system and method for generating an m-dimensional signature vector in a computing device is provided. The signature vector may be generated from a plurality of key-value pairs, each comprising a unique identifier and an associated non-zero value. Each element of the m-dimensional signature vector is calculated based on a summation of a plurality of terms. Each of the terms is calculated from a respective key-value pair by generating a seed based on the key of the respective key-value pair and an element identifier associated with the vector element being calculated; generating a pseudo-random number from the generated seed; and multiplying the pseudo-random number by the value of the respective key-value pair, wherein m<<n.

Abstract: A method for coding image data together with its shape and other details as a two dimensional barcode is disclosed. The method stores the shape details together with the color information of a product which is very useful for representing to a user. The method enables the encoding of an image with its shape, size, name and other details of the product in a most compact way. Also, the method enables decoding of the information to get the detailed information without using internet.

Abstract: A non-transitory computer readable storage medium stores one or more computer programs adapted to cause a processor based system to execute steps that include analyzing an image, identifying one or more faces in the image using a face recognition technique, designating at least one of the identified faces collectively as a first area of interest, and determining whether an insertion area exists in the image where additional content can be inserted without obstructing the first area of interest. Another computer program is adapted to cause a processor based system to execute steps that include determining whether the insertion area can be divided into two or more regions based on color. Methods and processor based apparatuses that perform one or more of these steps are also disclosed.

Abstract: According to one embodiment, an image identification apparatus comprises an image pickup unit, an illumination unit, an illumination control unit and an identification unit. The image pickup unit configured to pickup an image of an identified object. The illumination unit configured to irradiate light towards the image pickup area of the image pickup unit. The illumination control unit configured to change the irradiation condition of the illumination unit in accordance with the image pickup timing of the image pickup unit. The identification unit configured to identify the identified object according to the image picked-up by the image pickup unit.

Abstract: The present disclosure concerns methods and/or systems for processing, detecting and/or notifying for the presence of anomalies or infrequent events from data. Some of the disclose methods and/or systems may be used on large-scale data sets. Certain applications are directed to analyzing sensor surveillance records to identify aberrant behavior. The sensor data may be from a number of sensor types including video and/or audio. Certain applications are directed to methods and/or systems that use compressive sensing. Certain applications may be performed in substantially real time.

Abstract: A method and gem pattern matching technique to analyze a target gemstone by analyzing a pattern created by transmitting a light source such as a laser beam through the gemstone to create a visual optical pattern and comparing the pattern to a database of known gemstone patterns to determine the percentage likelihood that the target gemstone will match a gemstone in the database. The matching is based on the weight of the heaviest spot in the pattern and its location in the gemstone image and comparing it to the weight and location of the heaviest spots in each gemstone image in the database to determine a percentage matching.

Abstract: A method for classifying a document (3) to be associated with at least one service (Si), including a step in which a scanner (1) having a processor (6) scans (E1) the document (3). The method also includes steps in which the processor (6): develops (E2) at least one structure (?j) representing the document (3), determines (E3) for each service (Si) at least one similitude value (?ij) between the structure (?j) representing the document and a reference structure (Rij) of the same kind and representing the service (Si), deduces (E4) from the similitude value (?ij) the service (Si) with which the document (3) is to be associated, and processes (E5) the document (3) according to the service (Si) thus associated. The invention also relates to a scanner for implementing the method.

Abstract: Techniques for identifying documents sharing common underlying structures in a large collection of documents and processing the documents using the identified structures are disclosed. Images of the document collection are processed to detect occurrences of a predetermined set of image features that are common or similar among forms. The images are then indexed in an image index based on the detected image features. A graph of nodes is built. Nodes in the graph represent images and are connected to nodes representing similar document images by edges. Documents sharing common underlying structures are identified by gathering strongly inter-connected nodes in the graph. The identified documents are processed based at least in part on the resulting clusters.

Abstract: A digital image of a first garment having one or more first garment portions is received. A user has identified the first garment portions as matching one or more corresponding second garment portions of a second garment. The probability of accidental match of the first garment within the digital image in relation to the second garment is determined, by using a statistical model based on one or more parameters and based on analysis of the first garment portions. The probability of accidental match is output.

Abstract: A device/system and method for creating customized audio segments related to an object of interest are disclosed. The device and/or system can create an additional level of interaction with the object of interest by creating customized audio segments based on the identity of the object of interest and/or the user's interaction with the object of interest. Thus, the mobile device can create an interactive environment for a user interacting with an otherwise inanimate object.

Abstract: A system and a method that facilitates generating a model from a 3-dimensional (3D) object assembled from 2-dimensional (2D) content are described. A content aggregator can construct a 3D object from a collection of two or more 2D images each depicting a real entity in a physical real world, wherein the 3D object is constructed by combining the two or more 2D images based upon a respective image perspective. A 3D virtual environment can allow exploration of the 3D object. A model component can extrapolate a true 3D geometric model from the 3D object, wherein the true 3D geometric model is generated to include scaling in proportion to a size within the physical real world.

Abstract: A character recognition apparatus includes an acquisition unit, a specification unit, a movement unit, and a recognition unit. The acquisition unit acquires data representing a character string. The specification unit specifies an element of a compound character satisfying a predetermined condition for determining the compound character from the character string. The movement unit moves the element of the compound character close to an adjacent character. The recognition unit recognizes a changed character string in which the movement unit has moved the element of the compound character, based on a shape of characters and relevance between adjacent characters.

Abstract: Map data are overlaid on satellite imagery. A road segment within the map data is identified, and the satellite imagery indicates that the road segment is at a different geographic position than a geographic position indicated by the map data. The endpoints of the road segment in the map data are aligned with the corresponding positions of the endpoints in the satellite imagery. A road template is applied at an endpoint of the road segment in the satellite imagery, and the angle of the road template that matches the angle of the road segment indicated by the satellite imagery is determined by optimizing a cost function. The road template is iteratively shifted along the road segment in the satellite imagery. The geographic position of the road segment within the map data is updated responsive to the positions and angles of the road template.

Abstract: A method of converting a raster scan image to a vector-based image including (a) reading a scanline of a raster scan image containing n pixels per scanline; (b) classifying n+1 groupings of four adjacent pixels within the scanline using only the four adjacent pixel values; (c) assigning to each grouping (x,y), for all groupings in the scanline, plural vectors based on (1) the classification of the grouping (x,y) and (2) vectors assigned to a grouping (x?1,y) and vectors assigned to a grouping (x, y?1); (d) adding a point (x,y) to at least one vector assigned to the grouping (x,y) when the classification of grouping (x,y) indicates that the point (x,y) corresponds to a corner; (e) determining whether the points added to any vector form a closed loop; (f) writing out the points added to any vector that form a closed loop; and (g) repeating steps (a)-(f) for all of the scanlines of the raster scan image.

Abstract: Techniques to manage a whiteboard for multimedia conference events are described. An apparatus may comprise a whiteboard manager component operative to manage whiteboard recording, image processing and reproduction operations for a multimedia conference event. The whiteboard manager component may comprise an image quantizer module operative to receive an image of a writing surface with pen strokes, and quantize each pixel of the image into a predetermined number of colors, an attribute extractor module communicatively coupled to the image quantizer module, the image quantizer module operative to extract stroke attribute information for the pen strokes from the quantized pixels, and a whiteboard interface module communicatively coupled to the attribute extractor module, the whiteboard interface module operative to send the stroke attribute information to a meeting console for rendering as digital pen strokes on a digital writing surface. Other embodiments are described and claimed.

Abstract: A computer-readable storage medium storing an image processing program that causes a computer to execute a process includes acquiring the same object identification information included in a plurality of image data items by referring to a storage unit that stores each of the image data items, object identification information that identifies an object included in the image data item, and a location information item that identifies a location of the object in the image data item in association with one another; acquiring the location of the object identified by the acquired object identification information in each of the image data items by referring to the storage unit; computing a difference between the acquired location information items; comparing the difference between the location information items with a predetermined location threshold value; and determining whether the image data items are to be in the same group.

Abstract: A system is proposed for automated document processing, comprising a document, consisting of two sections—a main section, containing data in printed character form, and a supplementary section in a machine-readable form; a document forming means; a document inputting means; a character recognition means; a main and supplementary data comparison means. Said system uses the supplementary section data to confirm the main section data. The supplementary section data can fully or partly duplicate the main section data, supplement it and also comprise other additional data. The supplementary machine-readable section can be realized in a form of coded consecutive characters, printed graphic image (bar-code), magnetic, optical, microprocessor or other kind of data storage means. For enhancing security of documents all or a part of data can be coded prior to introduction into the supplementary section.

Abstract: Described herein is a framework for automatically classifying a structure in digital image data are described herein. In one implementation, a first set of features is extracted from digital image data, and used to learn a discriminative model. The discriminative model may be associated with at least one conditional probability of a class label given an image data observation Based on the conditional probability, at least one likelihood measure of the structure co-occurring with another structure in the same sub-volume of the digital image data is determined. A second set of features may then be extracted from the likelihood measure.

Abstract: A method is provided for training and using an object classifier to identify a class object from a captured image. A plurality of still images is obtained from training data and a feature generation technique is applied to the plurality of still images for identifying candidate features from each respective image. A subset of features is selected from the candidate features using a similarity comparison technique. Identifying candidate features and selecting a subset of features is iteratively repeated a predetermined number of times for generating a trained object classifier. An image is captured from an image capture device. Features are classified in the captured image using the trained object classifier. A determination is made whether the image contains a class object based on the trained object classifier associating an identified feature in the image with the class object.

Abstract: An image data generating device includes an original image acquirer that acquires original image data of a subject. The image data generating device also includes a controller that generates image data. The controller includes a monochrome image data generator that generates monochrome image data from the original image data. The controller also includes a color image data generator that generates color image data from the original image data. Further, the controller includes a text data acquirer that acquires recognition result data from the monochrome image data generated by the monochrome image data generator. In addition, the controller includes a recording unit that associates and records the color image data generated by the color image data generator and the recognition result data acquired by the text data acquirer.

Abstract: The present invention discloses a dual-vision driving safety warning device and a method thereof. The device of the present invention comprises an image capture unit, an image processing unit, a vehicle status sensing unit, a warning judgment logic, and at least one warning unit. The image capture unit includes at least two image capture devices installed in a user's vehicle and capturing the images of the front traffic environment. The image processing unit processes the images and uses the vehicle status signals detected by the vehicle status sensing unit to calculate the distance between the user's vehicle and a front vehicle in the image. The warning judgment logic sends out a control signal when the user's vehicle deviates from a driving lane or approaches a front vehicle too much. The warning unit receives the control signal and sends out a warning signal, such as a sound or a flash, to remind the driver.

Abstract: A method for processing image data comprises generating face data representing a set of detected faces from image data representing a set of images forming a library of images, using the face data to determine a sub-set of the images from the library in which at least the same person appears, and generating at least one clothing signature for the person representing clothes worn by the person in at least one of the images in the sub-set, and using the or each clothing signature, identifying further images of the person in the library.

Abstract: The present invention relates to a method of identifying a target object in an image using image processing. It further relates to apparatus and computer software implementing the method. The method includes storing template data representing a template orientation field indicative of an orientation of each of a plurality of features of a template object; receiving image data representing the image; processing the image data to generate an image orientation field indicating an orientation corresponding to the plurality of image features; processing the image orientation field using the template orientation field to generate a match metric indicative of an extent of matching between at least part of the template orientation field and at least part of the image orientation field; and using the match metric to determine whether or not the target object has been identified in the image. Image and/or template confidence data is used to generate the match metric.

Abstract: Methods, apparatuses, and systems for image-based measurement and inspection of pre-engineered structural components, such as building trusses and wall panels. A system can include: a light source; a camera; a first memory storage; a second memory storage; and a processing unit configured to (i) detect a characteristic of the structural component, (ii) compare the characteristic to a corresponding characteristic of at least one reference data, and (iii) indicate a result of the comparison. A method can include: causing a light source to illuminate a portion of the structural component, receiving a reflection of the light source from the illuminated portion of the structural component, and storing data corresponding to the intensity of the reflection; comparing the stored data to at least one reference data; and indicating a result of the comparison.

Abstract: Registration and classification of non-textual information, such as digital images and video is described. Image searching and comparison of the images is also described. The digital images are indexed (i.e., each image is assigned a unique numerical parameter and/or a plurality of numerical parameters). The resulting index files are stored in a database that can be quickly searched because the index files are universal numerical files that are significantly smaller in size than their source images. Image search queries are also indexed to generate an index file, which is then compared with the stored index files. A similarity score is also calculated to rank the similar images based on the index file-to-index file comparison.

Abstract: A method for geolocating photographs is implemented in a computing device. The method includes receiving a photograph captured by a user as data stored on a computer-readable medium, receiving location history data for the user, determining a time at which the photograph was captured based on an object depicted in the photograph, determining a geographic location at which the photograph was captured using the determined time and the location history data.

Abstract: A method is provided for tracing a line from a possible corner of a pallet. The method may comprise: providing a Ro image; identifying, using a computer, a possible pallet corner in the Ro image; positioning, using the computer, a J×K window over at least respective portions of a plurality of rows in the Ro image including at least a portion of a row containing the possible corner; calculating, using the computer, an average of pixel values for each row in the J×K window; determining, using the computer, one of the rows in the J×K window having an average pixel value nearest a current pixel location being considered for inclusion in a line being traced and defining the one row as a nearest row; and deciding, using the computer, whether the nearest row is over a pallet.

Abstract: A method and apparatus to enable a position of an image to be identified and to enable the image to be recognized reliably even when the image is unclean or external disturbance light is irradiates the image. On a assumption that a binarization threshold value is calculated for every block based on the maximum and minimum values of the brightness values in code symbol data, plural adjacent blocks, the threshold values of which are calculated, are grouped to generate groups. The area of each of the groups thus generated is compared with a predetermined area for a comparison standard and if the area of each of the groups thus compared is larger than that, these groups are selected. The position of the selected groups is identified. Brightness value of the code symbol data are binarized using the threshold value for every block in the selected groups.

Abstract: Techniques for facilitating a shape-based search of content items are disclosed. In some situations, the techniques include receiving a request for search of content that includes an image of an item and a glyph input by the user on the image that indicates a shape of the first item, extracting features associated with the item from an area of the image on which the glyph is input; identifying a shape category of the first item based at least in part on the indicated shape of the glyph; and conducting a search for items having features that match the extracted features of the first item, the search being conducted among a plurality of items associated with the identified shape category. The retrieved search results may be returned to a user.

Abstract: A method determines similarity of objects depicted in images when the images pertain to different modalities. The method includes obtaining images that depict the objects and that pertain to the different modalities. An embedding function is applied to each of the images. The embedding function is selected from a set of two or more embedding functions, each of the embedding functions corresponding to a modality of the different modalities, the selected embedding function corresponding to the modality of the image to which it is applied. Application of the embedding function maps that image to a representation in a representation space such that when the images are mapped to the representation space, a distance between the representations of the images is indicative of a similarity of their depicted objects. The similarity of the depicted objects is determined based on the location of the corresponding representations in the representation space.

Abstract: According to one embodiment, a device for identifying types of document files includes an extraction unit, an evaluation unit, and a document type identification unit. The extraction unit extracts document feature information indicating distinguishing information for a document included in a document file. The evaluation unit verifies the validity of the document feature information extracted by the extraction unit. The document type identification unit identifies a type of the document file corresponding to the document feature information extracted by the extraction unit if the document feature information is valid.

Abstract: A method for training a pattern recognition algorithm including the steps of identifying the known location of the pattern that includes repeating elements within a fine resolution image, using the fine resolution image to train a model associated with the fine image, using the model to examine the fine image resolution image to generate a score space, examining the score space to identify a repeating pattern frequency, using a coarse image that is coarser than the finest image resolution image to train a model associated with the coarse image, using the model associated with the coarse image to examine the coarse image thereby generating a location error, comparing the location error to the repeating pattern frequency and determining if the coarse image resolution is suitable for locating the pattern within a fraction of one pitch of the repeating elements.

Abstract: An imaging apparatus configured to associate character information detected from an imaged picture displayed on a screen with a recorded picture displayed on the same screen including an imaging unit configured to image a subject and output an imaged picture, a picture recording unit configured to record pictures, a playback unit configured to play back a recorded picture recorded in the picture recording unit, and a character information detection unit configured to detect character information from the imaged picture outputted by the imaging unit.

Abstract: An image processing method, including extracting compensation information comprising one from among a depth compensation value and a depth value compensated for by using the depth compensation value; when the compensation information comprises the depth compensation value, compensating for a depth value to be applied to a pixel of a two-dimensional (2D) image by using the depth compensation value, and generating a depth map about the 2D image by using the compensated depth value, and when the compensation information comprises the compensated depth value, generating the depth map about the 2D image by using the compensated depth value; obtaining positions in a left-eye image and a right-eye image by using the depth map, wherein the pixel of the 2D image is mapped to the positions; and generating the left-eye image and the right-eye image comprising the positions to which the pixel is mapped.

Abstract: The present invention is directed toward a system and process that controls a group of networked electronic components using a multimodal integration scheme in which inputs from a speech recognition subsystem, gesture recognition subsystem employing a wireless pointing device and pointing analysis subsystem also employing the pointing device, are combined to determine what component a user wants to control and what control action is desired. In this multimodal integration scheme, the desired action concerning an electronic component is decomposed into a command and a referent pair. The referent can be identified using the pointing device to identify the component by pointing at the component or an object associated with it, by using speech recognition, or both. The command may be specified by pressing a button on the pointing device, by a gesture performed with the pointing device, by a speech recognition event, or by any combination of these inputs.

Abstract: The invention relates to a method for texturizing virtual three-dimensional objects, particularly virtual three-dimensional building objects and city models with a photographic image (1) of a real object, particularly of a picture of a façade.

Abstract: A digital video camera system that provides a video summary using a method that includes: designating a reference image containing a particular person; capturing a video sequence of the scene using the image sensor, the video sequence including a time sequence of image frames; processing the captured video sequence using a video processing path to form a digital video file; during the capturing of the video sequence, analyzing the captured image frames using a person recognition algorithm to identify a subset of the image frames that contain the particular person; forming the video summary including fewer than all of the image frames in the captured video sequence, wherein the video summary includes at least part of the identified subset of image frames containing the particular person; storing the digital video file in the storage memory; and storing a representation of the video summary in the storage memory.

Abstract: A labeling unit that extracts a character included in raster format image data and finds the size of the extracted character, and an approximation method determination unit that determines whether or not to use curve approximation to convert the character based on the found size, are provided. In the case where it has been determined to use curve approximation, the character is converted into a vector format by performing straight-line approximation or curve approximation on each of the various parts of the contour of the character, whereas in the case where it has been determined not to use curve approximation, the character is converted into a vector format by performing straight-line approximation and not performing curve approximation on each of the various parts of the contour of the character.

Abstract: A biometric information processing device includes, a biometric sensor configured to acquire a plurality of different biometric information elements; an authentication feature extracting unit configured to extract an authentication feature for use in authentication for each of the plurality of different biometric information elements acquired by the biometric sensor; a supplemental feature extracting unit configured to extract a supplemental feature relating to the authentication feature for each of the plurality of different biometric information elements; and a combined feature extracting unit configured to extract a combined feature in which a plurality of the supplemental features extracted by the supplemental feature extracting unit are reflected.

Abstract: A method, system and computer program product for matching images is provided. The images to be matched are represented by feature points and feature vectors and orientations associated with the feature points. First, putative correspondences are determined by using feature vectors. A subset of putative correspondences is selected and the topological equivalence of the subset is determined. The topologically equivalent subset of putative correspondences is used to establish a motion estimation model. An orientation consistency test is performed on the putative correspondences and the corresponding motion estimation transformation that is determined, to avoid an infeasible transformation. A coverage test is performed on the matches that satisfy orientation consistency test. The candidate matches that do not cover a significant portion of one of the images are rejected. The final match images are provided in the order of decreasing matching, in case of multiple images satisfying all the test requirements.

Abstract: One embodiment relates to a method of identifying sensitive expressions in images for a language with a large alphabet. The method is performed using a computer and includes (i) extracting an image from a message, (ii) extracting image character-blocks (i.e. normalized pixel graphs) from the image, and (iii) predicting characters to which the character-blocks correspond using a multi-class learning model, wherein the multi-class learning model is trained using a derived list of sensitive characters which is a subset of the large alphabet. In addition, (iv) the characters may be combined into string text, and (v) the string text may be searched for matches with a predefined list of sensitive expressions. Another embodiment relates to a method of training a multi-class learning model so that the model predicts characters to which image character-blocks correspond. Other embodiments, aspects and features are also disclosed herein.

Abstract: A method for identifying a stream of images, comprises: a) capturing one or more frames from the media which is to be identified; b) sending the selected image(s) as queries to a server; c) in the server, analyzing the one or more images to identify them from a database that for each movie stored comprises switched significantly different frames; d) if more than one image is processed, aggregating the various server's replies; and e) displaying the results to the user.

Abstract: A system and method are provided for controlling a device based on computer vision. Embodiments of the system and method of the invention are based on receiving a sequence of images of a field of view; detecting movement of at least one object in the images; applying a shape recognition algorithm on the at least one moving object; confirming that the object is a user hand by combining information from at least two images of the object; and tracking the object to control the device.

Abstract: A method and system for automatic detection and volumetric quantification of bone lesions in 3D medical images, such as 3D computed tomography (CT) volumes, is disclosed. Regions of interest corresponding to bone regions are detected in a 3D medical image. Bone lesions are detected in the regions of interest using a cascade of trained detectors. The cascade of trained detectors automatically detects lesion centers and then estimates lesion size in all three spatial axes. A hierarchical multi-scale approach is used to detect bone lesions using a cascade of detectors on multiple levels of a resolution pyramid of the 3D medical image.

Abstract: A fingerprint of an image identified within a received message is generated following analysis of the message. A spam detection engine identifies an image within a message and converts the image into a grey scale image. The spam detection engine analyzes the grey scale image and assigns a score. A fingerprint of the grey scale image is generated based on the score. The fingerprint may also be based on other factors such as the message sender's status (e.g. blacklisted or whitelisted) and other scores and reports generated by the spam detection engine. The fingerprint is then used to filter future incoming messages.

Abstract: A method is provided in one embodiment and includes establishing a communication session involving a first endpoint and a second endpoint that are associated with a video conference in a network environment. The first endpoint is associated with a first identifier and the second endpoint is associated with a second identifier. The method also includes evaluating first audio data for the first endpoint, and determining a vocative parameter associated with the first audio data, where image data can be rendered on a user interface at the first endpoint based on the detecting of the vocative parameter.

Abstract: The present invention provides an image recognition apparatus with enhanced performance and robustness. In an image recognition apparatus 1, an image classification information accumulating unit 20 stores therein feature information defining visual features of various objects obtained through a learning process. For classification of input images, an image feature obtaining unit 18 extracts descriptors indicating features from each input image, image vocabularies corresponding to the descriptors are voted, and a classifying unit 19 calculates existence probabilities of one or more candidate objects, based on the result of the voting. According to the existence probabilities, objects contained in the image is identified. Through the calculation, the existence probabilities are adjusted by an exclusive classifier, based on exclusive relationship information defining exclusive object sets each containing different objects (object labels) predicted not to coexist in a same image.

Abstract: According to one embodiment, a visual detection system includes an image processing unit coupled to a display and one or more neuro-physiological sensors. The display generates an image that may be viewed by a user. The neuro-physiological sensors generate neuro-physiological signals representing neuro-physiological characteristics of the user while viewing the display. The image processing unit receives the neuro-physiological signals from the neuro-physiological sensors, determines a region of interest in the image according to the neuro-physiological signals, and provides an indication of the presence of the region of interest in the image.