Abstract: A selection apparatus generates comparison information based on comparison between the pixel values of input pixels and a reference pixel value. The selection apparatus selects the output pixels from the input pixels based on the comparison information and feature amounts representing the features of the input pixels.

Abstract: A system and method of for describing target data as a sequence of pattern elements and pattern element groups that comprise an overall target pattern is described. Pattern elements may utilize regular expression syntax along with other metadata that describe the behavior of the element. A pattern element group may be a collection of fully defined pattern elements where at least one pattern element from the group must have a match for the overall pattern to match. Patterns contain both pattern elements and pattern element groups. The general process involves first performing optical character recognition (OCR) on the document, which in turn produces a sequence of text tokens representing the lines of text on each page of the document. The search algorithm may then apply each defined pattern to the entire document capturing and/or extracting data that match each pattern's required elements and element groups.

Abstract: Methods and apparatus for a roof analysis tool for constructing a parameter set, where the parameter set is derived from mapping data for a map region, and where the parameter set describes the roofs for the buildings within the map region. In some cases, the parameter set includes a list of roof type identification values and the respective buildings in the map region for which a given roof type identification value corresponds. The roof analysis tool may operate on a server and work in conjunction with a mobile device, where the mobile device may display map views of a map region such that the map view is based on a three-dimensional model of the map region, and where a portion of the three-dimensional model is based on data generated on the mobile device and a portion of the three-dimensional model is based on data generated on the server.

Abstract: Apparatuses and methods are provided for determining whether a user intends to provide an input using an image of a control appearing on a surface. An apparatus may include first and second cameras to capture set and second sets of two or more images of the surface, and a unit to determine whether various conditions are true. A first condition is that a particular number of skin-color pixels are present within one cell of the two or more images. A cell location substantially coincides with the image of the control. A second condition is that the skin-color pixels persist for at least a particular time period. A third condition is that a region of skin-color pixels has a substantially rectangular shape. A fourth condition is that corresponding edges of skin-color pixels in the first and second sets of images are within a predefined offset distance. Additional conditions are disclosed.

Abstract: Among other things, methods, systems and computer program products are disclosed for managing image files. In one aspect, a method includes displaying, at a data processing device, a preview of images available for import from a storage device, wherein each image is associated with a pair of corresponding image files comprising a compressed image file and a raw image file obtained from a single image capture. At the data processing device, both the raw image file and the compressed image file of the pair associated with at least one of the images are imported. At the data processing device, at least one of the pair of corresponding image files is identified as a master image file for the associated image. At the data processing device, the identified master image file for the associated captured image is displayed.

Abstract: Computerized appliances may be operated by users remotely. In one implementation, a learning controller apparatus may be operated to determine association between a user indication and an action by the appliance. The user indications, e.g., gestures, posture changes, audio signals may trigger an event associated with the controller. The event may be linked to a plurality of instructions configured to communicate a command to the appliance. The learning apparatus may receive sensory input conveying information about robot's state and environment (context). The sensory input may be used to determine the user indications. During operation, upon determine the indication using sensory input, the controller may cause execution of the respective instructions in order to trigger action by the appliance. Device animation methodology may enable users to operate computerized appliances using gestures, voice commands, posture changes, and/or other customized control elements.

Abstract: A method of switching a display device according to one embodiment of the present specification can include the steps of displaying an image received from a rear side camera of the display device on a touch sensitive display unit positioned at the front side of the display device, setting an area on which a predetermined gesture is to be recognized based on a touch input and if a gesture substantially identical to the predetermined gesture is recognized on the set area in the image received from the rear side camera, capturing an image using the rear side camera.

Abstract: A display device which may enhance recognition precision of an operation command input by a user, a mobile terminal which communicates with the display device, and a method of controlling the same are provided. The display device includes a display panel, an image collector configured to collect an image of an object in a first direction, a communicator configured to receive an image of the object collected in a second direction by a mobile terminal, and a controller configured to process the collected image and the received image together into a processed image, to recognize a motion of the processed image, determine an operation command corresponding to the recognized motion, and control an operation of the display panel based on the determined operation command.

Abstract: A reasoning engine is disclosed. Contemplated reasoning engines acquire data relating to one or more aspects of various environments. Inference engines within the reasoning engines review the acquire data, historical or current, to generate one or more hypotheses about how the aspects of the environments might be correlated, if at all. The reasoning engine can attempt to validate the hypotheses through controlling acquisition of the environment data.

Abstract: A method for visual recognition of an object in an electronic image includes extracting unique points of an object to be learned and/or a target object. The unique points are obtained by cross-correlating the image with a structure. Generally, the structure and/or the size of the structure may vary to detect extremum information associated with the learned object and/or target object. An icon corresponding to each of the unique points is extracted. The size of the icon corresponds to the scale of the unique point. After extraction of the various icons, an object becomes a collection of icons. Each of these icons is un-rotated and normalized or resized to a constant size so it can be compared with other icons. One of the unique properties of these icons is their stability over scale and angle. Thus, this invention allows the recognition of an image(s) or object(s) from large number of trained images or objects very quickly.

Abstract: Image segmentation utilizing the color and brightness of sections of the image to generate super-pixels is described herein. The image segmentation is implemented pixel-wise. A color triangle is utilized in the classification of the pixels. By using a hierarchical classification table, the memory usage is significantly reduced, and the image segmentation is able to be performed on-the-fly in real-time.

Abstract: An intra prediction mode determination apparatus includes prediction units which perform intra prediction on a block of a first block size in an image including at least one block, a first determination unit which determines one of the prediction units used to perform the intra prediction on the block of a processing target, and a second determination unit which determines an intra prediction mode for the block of the processing target from among candidates of an intra prediction mode which are used by the determined prediction unit in accordance with a result of the intra prediction performed by the determined prediction unit. When the block of the processing target includes a block of a second block size which is larger than the first block size, the prediction unit performs intra prediction on blocks of the first block size obtained by dividing the block of the second block size.

Abstract: The method of acquiring image comprises: acquiring a first image; determining a first sub-image corresponding to a first object from the first image; and replacing the first sub-image with a second sub-image different from the first sub-image to obtain a second image. This solution may replace the first sub-image with the second sub-image, thereby achieving the technical effect of automatically removing a part of content from the image during image acquiring.

Abstract: A method for compressing a bi-level document image containing text is disclosed. The document image is segmented into symbol images each representing a letter, numeral, etc. in the document. The symbol images are classified into a plurality of classes, each class being associated with a template image and a class index. Classification is done by comparing each symbol to be classified with template of existing classes, using a number of image features including zoning profiles, side profiles, topology statistics, and low-order image moments. These image features are compared using a tolerance based method to determine whether the symbol matches the template. After classification, certain classes that have few symbols classified into them may be merged with other classes. In addition, the template images of the classes are down-sampled, where the final sizes of the template images are dependent on the likelihood of confusion of the template with other templates.

Abstract: In a method for identifying and classifying an object, an object is detected by at least one physical detector tuned for it, the object is evaluated from the output signal of the detector and by an evaluation unit, and the object is identified and/or classified on the basis of predefinable properties from the output signal. A number of different physical features of the object are derived from the output signal, and the object is assigned to one of N predetermined basic classes on the basis of the derived physical features. The N basic classes are arranged in a predetermined order to form an N-dimensional vector V, which is assigned to the object, such that the elements v1, . . . , vN of the vector V indicate that the object belongs to the respective basic class. The object is then assigned to a derived class, which is taken from a reference data base, as a function of the vector V.

Abstract: An image acquisition section of an information processor acquires stereo images from an imaging device. An input information acquisition section acquires an instruction input from a user. A depth image acquisition portion of a position information generation section generates a depth image representing a position distribution of subjects existing in the field of view of the imaging device in the depth direction using stereo images. A matching portion adjusts the size of a reference template image in accordance with the position of each of the subjects in the depth direction represented by the depth image first, then performs template matching on the depth image, thus identifying the position of a target having a given shape and size in the three-dimensional space. An output information generation section generates output information by performing necessary processes based on the target position.

Abstract: Embodiments of a system and methods for skin detection and pose determination of a hand in an image are generally described herein. A method for may include detecting skin pixels in an image using a multi-layer skin filter and classifying the skin pixels into a set of foreground skin pixels and a set of background skin pixels. The method may include storing information about the set of background skin pixels in a persistent memory. The method may include determining a set of features from the set of foreground skin pixels, and clustering features from the set of features to create a set of hand pose descriptors. The method may include determining a set of candidate hand pose region descriptors and a set of candidate hand pose contour descriptors that match the set of hand pose descriptors and detecting a valid hand pose using the sets of candidate descriptors.

Abstract: There is provided an information processing device including an image obtaining unit configured to obtain an image including a plurality of objects, and a display control unit configured to, after an operation has been performed on a first object of the plurality of objects, control identification information associated with a second object, the identification information indicating the second object.

Abstract: A system and method that is able to recognize a user's natural superimposed handwriting without any explicit separation between characters. The system and method is able to process single-stroke and multi-stroke characters. It can also process cursive handwriting. Further, the method and system can determine the boundaries of input words either by the use of a specific user input gesture or by detecting the word boundaries based on language characteristics and properties. The system and method analyzes the handwriting input through the processes of segmentation, character recognition, and language modeling. These three processes occur concurrently through the use of dynamic programming.

Abstract: A method and an apparatus search an image using a feature point. The image search method extracts at least one feature point from an image and describes the extracted at least one feature point in stages, thereby generating a hierarchical feature point descriptor. In addition, the method may search for information matching the feature point descriptor from a local database (DB) included in a terminal or a remote DB included in a server.

Abstract: An information processing system that controls a display to display an image generated based on a manipulation target image and an image based on a manipulation target region that is set based on a detection of an object included in acquired image data.

Abstract: An extraction device includes a processor that executes a procedure. The procedure includes identifying, as a video data portion including a scene of interest in a sports game, a portion of, or all of, captured images of the sports game that correspond to a period in which audio data obtained by collecting sound from the sports game exhibits relatively loud sound, or a period in which the audio data exhibits sound louder than a specific loudness.

Abstract: Techniques related to labeling component parts and detecting component properties in imaging data are discussed. Such techniques may include generating a feature vector including invariant features associated with an area of interest within an image of an object such as an image of a hand and providing a component label such as a hand part label for the area of interest based on an application of a machine learning classifier to the feature vector.

Abstract: Methods, systems and articles of manufacture for recognizing and locating one or more objects in a scene are disclosed. An image and/or video of the scene are captured. Using audio recorded at the scene, an object search of the captured scene is narrowed down. For example, the direction of arrival (DOA) of a sound can be determined and used to limit the search area in a captured image/video. In another example, keypoint signatures may be selected based on types of sounds identified in the recorded audio. A keypoint signature corresponds to a particular object that the system is configured to recognize. Objects in the scene may then be recognized using a shift invariant feature transform (SIFT) analysis comparing keypoints identified in the captured scene to the selected keypoint signatures.

Abstract: Generally, an apparatus and a method for encoding multiview video downscale two views of the multiview video, determine a shift between the downscaled views, and encode one the two views using the shift.

Abstract: A method for enhancing fast backward performance includes: with regard to a plurality of offsets of a multimedia file, respectively storing corresponding cluster numbers into a first buffering region/buffer, where the offsets respectively correspond to different playback time points, and the cluster numbers respectively represent a plurality of clusters belonging to the multimedia file; and utilizing at least one portion of the offsets and the cluster numbers to perform a fast backward operation of the multimedia file. An associated electronic device is further provided.

Abstract: Speech recognition systems and/or techniques are provided in which grammar elements and/or speech inputs are targeted to selected functions. One or more input capture devices facilitate the collection of user input associated with a vehicle, and a vehicle function may be selected based upon received user input. A subset of available grammar elements that are associated with audible commands for the selected function may then be identified and utilized to evaluate received audio input. In this regard, speech recognition may be targeted to the selected function.

Abstract: An object identification system comprises at least one database of reference images. Each reference image comprises an identifiable reference object and is associated in the database with a corresponding text description of the reference object. The system also comprises image query resolution data structures, each containing a collection of records having an image descriptor of one of the reference images, and searchable using a corresponding search algorithm. Upon receiving a query comprising an input image including a query object, the system generates a query image descriptor of the input image corresponding with one or more search algorithms, and searches the corresponding image query resolution data structure to identify a closest match record. A best available match record is identified from the closest match records identified in the image query resolution data structures, and a text description of the query object is retrieved corresponding with the associated reference image.

Abstract: Financial transaction card data can be entered by providing a picture of the card to a server programmed with a text recognition algorithm. The server can perform text recognition on the image at the same time that a consumer enters additional required data, such as a zip code. The server can perform as much text recognition processing as possible in the time the consumer is entering the additional data. Once the additional data is received, a signal can be provided to the server indicating that the user is now waiting for results of the text recognition process, meaning the server should provide them as quickly as possible. Once text recognition results are received, a consumer can make a selection to identify a character which the text recognition algorithm did not sufficiently identify. Based on known account number constraints, the user selection can cause multiple characters to be identified.

Abstract: An example collation apparatus includes a user information storage medium and processing circuitry. The processing circuitry is configured to at least receive a captured image including a face of a user and a non-facial object; detect a face region from the image; set an object region corresponding to the location of the non-facial object in the image; extract a face feature vector from the face region and collate the extracted face feature vector with face feature vectors stored in the user information storage medium; extract an object feature vector from the object region and collate the extracted object feature vector with object feature vectors stored in the user information storage medium; and calculate a collation result of the user using at least one of a collation result of face feature vectors and a collation result of object feature vectors.

Abstract: One or more embodiments of a digital content system allow a user to conveniently search and/or navigate through a collection of digital content items. In particular, a user can interact with a client device to search for and identify one or more digital content items within a collection of digital content items. For example, the digital content system may provide a photo from a collection of photos via a graphical user interface. The digital content system can receive a user input identifying a selection of one or more visual features within the photo. Based on the selected visual feature(s), the digital content system may identify photos from the collection of photos that include the identified visual feature(s) and provide access to the identified photos via the graphical user interface.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time.

Abstract: A system and method for object instance localization in an image are disclosed. In the method, keypoints are detected in a target image and candidate regions are detected by matching the detected keypoints to keypoints detected in a set of reference images. Similarity measures between global descriptors computed for the located candidate regions and global descriptors for the reference images are computed and labels are assigned to at least some of the candidate regions based on the computed similarity measures. Performing the region detection based on keypoint matching while performing the labeling based on global descriptors improves object instance detection.

Abstract: The present technique has a face detection unit configured to detect a face region which is determined as a face in an image, a position specifying unit configured to specify a particular position in the image, a range specifying unit configured to set a hue range which has a first range containing a hue at a position specified by the position specifying unit in a case where the position specified by the position specifying unit is not contained in the face region and to set a hue range which has a second range containing a hue at a position specified by the position specifying unit and wider than the first range in a case where the position specified by the position specifying unit contains the face region, and image processing unit for performing image conversion processing based on color information contained in the hue range.

Abstract: Methods and systems to evaluate the ambient light in a local viewing environment where video content is being displayed, and to make adjustments to the ambient light, video display device, and/or video content to cause the perception of the color and brightness of the video content to the viewer to be true to what the video content producer intended the viewer to perceive.

Abstract: A neural network, wherein a portion of the neural network comprises: a first array having a first number of neurons, wherein the dendrite of each neuron of the first array is provided for receiving an input signal indicating that a measured parameter gets closer to a predetermined value assigned to said neuron; and a second array having a second number of neurons, wherein the second number is smaller than the first number, the dendrite of each neuron of the second array forming an excitatory STDP synapse with the axon of a plurality of neurons of the first array; the dendrite of each neuron of the second array forming an excitatory STDP synapse with the axon of neighboring neurons of the second array.

Abstract: Neural networks for object detection in images are used with a spatial pyramid pooling (SPP) layer. Using the SPP network structure, a fixed-length representation is generated regardless of image size and scale. The feature maps are computed from the entire image once, and the features are pooled in arbitrary regions (sub-images) to generate fixed-length representations for training the detectors. Thus, repeated computation of the convolutional features is avoided while accuracy is enhanced.

Abstract: An efficient image retrieval method utilizing topological order distance (TOD) algorithm is disclosed. In one aspect, an image recognition method may include steps of receiving one or more input images; extracting one or more key features on the input images; determining an origin and one or more feature points on the key features on the input images; establishing a topological relationship in each dimension between the feature points on the key features; and utilizing the topological relationship in each dimension between the feature points to identify the input images in an image database. Comparing with the conventional image retrieval techniques, the TOD algorithm is advantageous because it is more efficient, scale and rotation invariant, and tolerant to affine and perspective transformation.

Abstract: An electronic device is provided. The electronic device includes a memory configured to store a plurality of digital images and a processor, wherein the processor tracks movement of an object recognized in the plurality of digital images, and the processor detects an amount of movement between the plurality of digital images, and selects one among a plurality of object tracking methods based on at least a part of the amount of the movement.

Abstract: Provided is a device and method for estimating a head pose which may obtain an excellent head pose recognition result free from the influence of an illumination change, the device including a head area extracting unit to extract a head area from an input depth image, a head pitch angle estimating unit to estimate a pitch angle of a head in the head area, a head yaw angle estimating unit to estimate a yaw angle of the head in the head area, and a head pose displaying unit to display a head pose based on the estimated pitch angle of the head and the estimated yaw angle of the head.

Abstract: A method of enhancing the perceived sharpness of an image by controlling a spatial interpolation process. A pixel value spatial gradient measure is formed and used to generate a map of modified pixel positions. In this map, the pixel spacing is reduced when the gradient measure is high. By using this map of modified pixel positions to control an otherwise conventional interpolation process, an output image is formed having a perceived sharpness which is greater than that of the unmodified output of the interpolation process.

Abstract: Improved pattern detection is provided for a predefined pattern in data. A detection threshold employed by a Euclidean detector to detect a pattern (such as a Servo Address Mark) may be adjusted based upon data gathered that corresponds to an indication of a detected pattern by one of a plurality of pattern detectors.

Abstract: Systems, methods and computer readable media are disclosed for gesture shortcuts. A user's movement or body position is captured by a capture device of a system, and is used as input to control the system. For a system-recognized gesture, there may be a full version of the gesture and a shortcut of the gesture. Where the system recognizes that either the full version of the gesture or the shortcut of the gesture has been performed, it sends an indication that the system-recognized gesture was observed to a corresponding application. Where the shortcut comprises a subset of the full version of the gesture, and both the shortcut and the full version of the gesture are recognized as the user performs the full version of the gesture, the system recognizes that only a single performance of the gesture has occurred, and indicates to the application as such.

Abstract: A method for identifying and extracting a linear object from an image is disclosed. The method comprises: acquiring an original image to be processed, wherein the original image is taken by a camera, received through network transmission or copied from a compact disc or a removable disk; preprocessing the original image to obtain an enhanced image; extracting an edge information image from the enhanced image; then extracting linear features by performing, on the edge information image, a linear feature extracting transform improved with a cluster operator; finally; identifying and extracting the linear object by distinguishing the linear object from other linear features by considering characteristics of the linear object to be identified and extracted.

Abstract: Systems and methods are disclosed for determining whether a third party observer could determine that an organization has an intent with respect to subject matter based on the organization's web activity. The determination that there is a risk of information leaks to the third party observer can be completed by analyzing the entropy of web usage information destined for the third party observer's servers. Systems and methods are also disclosed for mitigating the risk of information leaks by obscuring the organization's web activity. The web activity can be obscured by selecting candidate actions that can be used to generate neutralizing web traffic from the organization's network which will obscure an intent with respect to a particular subject matter. For example, the candidate actions can identify specific queries, links, or actions that the organization can take to neutralize their web activity to a less remarkable point in the search space.

Abstract: An information retrieval system for automatically retrieving information related to the context of an active task being manipulated by a user. The system observes the operation of the active task and user interactions, and utilizes predetermined criteria to generate context representation of the active task that are relevant to the context of the active task. The information retrieval system then processes the context representation to generate queries or search terms for conducting an information search. The information retrieval system determines the relevance of a word to the context by utilizing an adaptive weighting system. The information retrieval system assigns varying weights to different attributes of a word and calculates an accumulated weight of the word by accumulating all weights assigned to the word. The attributes may include word size, style, location of the word, etc.

Abstract: In an embodiment, an image processing device includes: a first calculating unit; a selecting unit; an extracting unit; a second calculating unit; a determining unit; and a generating unit. The first calculating unit calculates posture information of a subject from a first image including the subject. The selecting unit selects a second image associated with the posture information from a plurality of second images including clothing. The extracting unit extracts a first feature of the subject from a first contour in the first image and extracts a second feature from a second contour in the second image. The second calculating unit calculates a first difference between the first feature and the second feature. The determining unit determines a transparency that is higher as the first difference is larger. The generating unit generates a composite image combining the first image and the second image with the determined transparency.

Abstract: A low-cost, low-demand (physical and cognitive) interface for puppeteers that allows them to easily participate in the control of verbal and non-verbal activities of a set of virtual characters or avatars. The invention further provides a low-cost, unencumbered interface for participants that allows them to employ both natural movement and also verbal and non-verbal interaction with virtual characters. A network protocol supports real-time remote interaction even when dealing with relatively poor network connections; and an integrated after-action review system supports trainers in the processes of tagging and commenting on events, subsequently using these to assist reflection on the part of participants.

Abstract: Methods, apparatus, and articles of manufacture to measure geographical features using an image of a geographical location are disclosed. An example method includes identifying first features of a first open market from a first image of a first geographic area, identifying second features of a second image of a second geographic area, comparing the first features to the second features, and identifying a second open market in the second image based on the comparison.

Abstract: An image processing apparatus includes a computer configured to execute a program to perform an image processing including: creating difference data between first encoded form data and a form format corresponding to the form data, the first encoded form data being acquired by encoding form data according to a first compression parameter; decoding the created difference data; performing character recognition on each of the form data and the decoded difference data; comparing a first result of the character recognition performed on the form data and a second result of the character recognition performed on the decoded difference data, and determining whether the first result and the second result coincide with each other; and outputting the difference data based on a result of the determination of coincidence/noncoincidence of the first result and the second result.