Abstract: Techniques and structures are disclosed in which one or more distortion categories are identified for an image or video, and a quality of the image or video is determined based on the one or more distortion categories. The image or video may be of a natural scene, and may be of unknown provenance. Identifying a distortion category and/or determining a quality may be performed without any corresponding reference (e.g., undistorted) image or video. Identifying a distortion category may be performed using a distortion classifier. Quality may be determined with respect to a plurality of human opinion scores that correspond to a particular distortion category to which an image or video of unknown provenance is identified as belonging. Various statistical methods may be used in performing said identifying and said determining, including use of generalized Gaussian distribution density models and natural scene statistics.

Abstract: Embodiments include a scene classification system and method. In one embodiment, a method includes forming a first plurality of image features from an input image, processing the first plurality of image features in the first scene classifier.

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 bit code converter transforms a learning feature vector using a transformation matrix updated by a transformation matrix update unit, and converts the transformed learning feature vector into a bit code. When the transformation matrix update unit substitutes a substitution candidate for an element of the transformation matrix, a cost function calculator fixes the substitution candidate that minimizes a cost function as the element. The transformation matrix update unit selects the element while sequentially changing the elements, and the cost function calculator fixes the selected element every time the transformation matrix update unit selects the element, thereby finally fixing the optimum transformation matrix.

Abstract: Various examples are disclosed herein that relate to staged element classification. For example, one disclosed example provides a method of classifying elements by forming elements for classification into a plurality of first-level sets in a first stage, generating primary groups within the first-level sets based on element similarity, forming a plurality of second-level sets from the first-level sets in a second stage, generating secondary groups within the second-level sets based on element similarity, and merging a plurality of the primary and/or secondary groups based on element similarity.

Abstract: A technique for determining a characteristic of a face or certain other object within a scene captured in a digital image including acquiring an image and applying a linear texture model that is constructed based on a training data set and that includes a class of objects including a first subset of model components that exhibit a dependency on directional lighting variations and a second subset of model components which are independent of directional lighting variations. A fit of the model to the face or certain other object is obtained including adjusting one or more individual values of one or more of the model components of the linear texture model. Based on the obtained fit of the model to the face or certain other object in the scene, a characteristic of the face or certain other object is determined.

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: A technique for determining a characteristic of a face or certain other object within a scene captured in a digital image including acquiring an image and applying a linear texture model that is constructed based on a training data set and that includes a class of objects including a first subset of model components that exhibit a dependency on directional lighting variations and a second subset of model components which are independent of directional lighting variations. A fit of the model to the face or certain other object is obtained including adjusting one or more individual values of one or more of the model components of the linear texture model. Based on the obtained fit of the model to the face or certain other object in the scene, a characteristic of the face or certain other object is determined.

Abstract: An object detection method and system for detecting an object in an image utilizing an adaptive image scanning strategy is disclosed herein. An initial rough shift can be determined based on the size of a scanning window and the image can be scanned continuously for several detections of similar sizes using the rough shift. The scanning window can be classified with respect to a cascade of homogenous classification functions covering one or more features of the object. The size and scanning direction of the scanning window can be adaptively changed depending on the probability of the object occurrence in accordance with scan acceleration. The object can be detected by an object detector and can be localized with higher precision and accuracy.

Abstract: An automated system and a method for extracting a region of interest from a digital image are disclosed. The method includes identifying a subset of training images from a larger set of training images, each training image in the set having a respective identified region of interest. The subset is identified based on a measure of similarity between the digital image and the images in the set of training images. At least one region of interest is extracted from the digital image based on an analysis of the identified regions of interest in the subset of training images.

Abstract: An apparatus for determining a fake image includes an image-acquiring block for acquiring an image captured by and input from a camera; and a background-learning block for learning a background of the image to create a learning background. Further, the apparatus for determining the fake image includes a face extracting-block for extracting a face region of a person to be authenticated when an input image for authentication is transmitted from the camera; and an inter-background comparing block for comparing a present background of an input image with the learning background. Furthermore, the apparatus for determining the fake image includes a motion information-comparing block for extracting motion information in the face region and the present background to compare the same with each other; and a fake image-determining block for determining whether the input image is faked using the compared results of the motion information and the backgrounds.

Abstract: A multi-stage method of visual object detection is disclosed. The method was originally designed to detect humans in specific poses, but is applicable to generic detection of any object. A first stage comprises acts of searching for members of a predetermined general-class of objects (such as humans) in an image using a cognitive swarm, detecting members of the general-class of objects in the image, and selecting regions of the image containing detected members of the general-class of objects. A second stage comprises acts of searching for members of a predetermined specific-class of objects (such as humans in a certain pose) within the selected regions of the image using a cognitive swarm, detecting members of the specific-class of objects within the selected regions of the image, and outputting the locations of detected objects to an operator display and optionally to an automatic response system.

Abstract: Described is a system for visual object recognition using heterogeneous classifier cascades. Visual object recognition is one of the most critical tasks for video and image analysis applications. The present invention utilizes a cascade of classifiers, wherein each stage is dedicated to a certain task such as achieving high accuracy or reducing false alarms. The stages are then appropriately trained using either the training data or false alarm datasets, respectively. Additionally, the features that are employed by the classifier cascades are heterogeneous and complementary in that several types of features may be used. The system described herein has multiple applications in a variety of fields including automotive safety, factory automation, surveillance, force protection, and automatic target recognition.

Abstract: A plurality of features determined from at least a portion of an image containing information about an object are processed with an inclusive neural network, and with a plurality of exclusive neural networks, so as to provide a plurality of inclusive probability values representing probabilities that the portion of the image corresponds to at least one of at least two different classes of objects, and for each exclusive neural network, so as to provide first and second exclusive probability values representing probabilities that the portion of the image respectively corresponds. or not. to at least one class of objects. The plurality of inclusive probability values, and the first and second exclusive probability values from each of the exclusive neural networks, provide for identifying whether the portion of the image corresponds, or not, to any of the at least two different classes of objects.

Abstract: A level set tree feature detection machine is disclosed along with a method for detecting a level set tree feature. At least one pixilated image is provided. An electronic model is generated of the pixilated images. Maximal meaningful nodes for the pixilated images are determined.

Abstract: A face illumination normalization method includes acquiring a digital image including a face that appears to be illuminated unevenly. One or more uneven illumination classifier programs are applied to the face data to determine the presence of the face within the digital image and/or the uneven illumination condition of the face. The uneven illumination condition may be corrected to thereby generate a corrected face image appearing to have more uniform illumination, for example, to enhance face recognition.

Abstract: Technologies are generally described for methods and systems for generating a representation of a skin condition. The method may include receiving first image data relating to a first image of a skin condition and second image data relating to a second image of the skin condition. The method may further include performing a component analysis of the first and second image data to generate a first and second weighted component analysis results for the first and second image data. The weighted component analysis results may represent the respective image data as a weighted sum of components. The method may further include performing a sampling analysis to the weighted component analysis results to generate a weighted image representation for the skin condition. The weighted image representation may represent the skin condition as a weighted sum of the components.

Abstract: A pattern recognition process, a non-transitory computer program product, and a mobile terminal for pattern recognition. An input pattern is normalized and a reliable pattern is generated from the normalized pattern by using at least one morphological operator. The distance between the reliable pattern and selected templates which are selected from a template library using a decision tree, is calculated. The reliable patterns are classified into at least one of the classes of the selected templates by at least one non-parametric classification method.

Abstract: A method, system and computer product for visualizing affinities between objects. The method includes the steps of: forming a representation of a minimum spanning tree where the minimum spanning tree connects the plurality of objects based on a pairwise distance between the plurality of objects; forming a hierarchical cluster of the plurality of objects where the hierarchical cluster includes a level; agglomerating the plurality of objects based on the pairwise distance; displaying a view of the representation of the minimum spanning tree in a graphical user interface; receiving a user selection of a parameter containing a hierarchical level; and identifying, in the view, a target cluster that corresponds to the hierarchical level; where at least one of the steps is carried out using a computer device so that affinities between the plurality of objects are visualized.

Abstract: Systems, methods, and apparatus, including software tangibly stored on a computer readable medium, involve image processing. An image processing graph includes multiple nodes that each correspond to an image processing operation. Input image data relating to a particular input image is received. Region of interest data defining an output image region to be generated based on the input image data is received. A domain of defined node output data is identified for each of the nodes using the input image data, and the graph is traversed in a first traversal order. An identification of the first traversal order is stored. A region of requested node output data is identified for each of the nodes using the region of interest data, and the graph is traversed in a second traversal order. The second traversal order is based at least in part on the stored identification of the first traversal order.

Abstract: An image processing apparatus that detects a face area from image information includes a pattern database, a face-direction determining unit, and a face-area detector. The face-direction determining unit determines, as a face direction, a direction in which a face image in image information is upright based on determination pattern information stored in the pattern database. The face-area detector matches the direction of the image information with the direction of the determination pattern information based on the face direction, and detects a face area from the image information.

Abstract: A system and method for semantic event detection in digital image content records is provided in which an event-level “Bag-of-Features” (BOF) representation is used to model events, and generic semantic events are detected in a concept space instead of an original low-level visual feature space based on the BOF representation.

Abstract: There is provided an information processing apparatus including a multi-stage determining unit that includes determinators which each function as a node of an N-level tree structure (N is an integer value of 2 or more) in order to perform determination for classifying a determination target into at least one of a plurality of ranges. Each determinator performs determination of classifying the determination target into any one of two ranges, and the two ranges determined in each determinator include an overlapping portion. The present technology can be applied to an information processing apparatus that classifies data.

Abstract: Various technologies and techniques are disclosed that improve handwriting recognition operations. Handwritten input is received in training mode and run through several base recognizers to generate several alternate lists. The alternate lists are unioned together into a combined alternate list. If the correct result is in the combined list, each correct/incorrect alternate pair is used to generate training patterns. The weights associated with the alternate pairs are stored. At runtime, the combined alternate list is generated just as training time. The trained comparator-net can be used to compare any two alternates in the combined list. A template matching base recognizer is used with one or more neural network base recognizers to improve recognition operations. The system provides comparator-net and reorder-net processes trained on print and cursive data, and ones that have been trained on cursive-only data. The respective comparator-net and reorder-net processes are used accordingly.

Abstract: A plurality of images inputted in an image signal input portion are divided into a plurality of regions by an image dividing portion, and a feature value in each of the plurality of regions is calculated by a feature value calculation portion and divided into a plurality of subsets by a subset generation portion. On the other hand, a cluster classifying portion classifies a plurality of clusters generated in a feature space into any one of a plurality of classes on the basis of the feature value and occurrence frequency of the feature value. And a classification criterion calculation portion calculates a criterion of classification for classifying images included in one subset on the basis of a distribution state of the feature value in the feature space of each of the images included in the one subset.

Abstract: The present disclosure relates to the development of terrain image analogy filters which may be used to convert visible terrain imagery from a first condition to a second condition. For example, from a seasonal “daytime/summer” view to a “nighttime/winter” appearance. The image analogy filters may be generated using processors such as one or more graphics processing units.

Abstract: An object image detection device is disclosed that is able to rapidly detect an object image from an input image. The object image detection device includes an image block generation unit to generate plural image blocks from the input image for detecting the object images, an image classification unit to determine whether each of the image blocks includes one or more of the object images by using one or more features of the object images, and acquire the image blocks including the object images to be object image candidates; and a detection unit to sequentially detect the object images from the object image candidates. The image classification unit acquires the object image candidates based on a relative positional relationship between the image blocks and already-acquired object image candidates.

Abstract: A method is provided for detection of an airborne contaminant in a atmospheric environment. The method comprises capturing an air sample from the atmospheric environment; separating candidate particles of interest from particles of non interest in the air sample; generating an image of the candidate particles; identifying a contaminant from among the candidate particles by comparing the image of candidate particles with a plurality of stored reference images, each of which reflects a respective identified contaminant; and notifying a remote third party in response to detecting a contaminant from among the candidate particles.

Abstract: A method and apparatus of tile-based belief propagation are disclosed. An image is split into a number of tiles. Messages are iteratively generated within each of the tiles based on the messages from neighboring pixels to the tile at a previous iteration, wherein each message represents information of a state of the pixel. The generated messages for sending out of the tiles are stored. Labels are then determined based on the stored messages, wherein each label represents the state of the pixel.

Abstract: In an image classification method, dividing an input image into blocks; obtaining block features of each block of the image; performing an evaluation of each block based on the block features thereof; obtaining image features based on the evaluations of the blocks of the image; and classifying the image based on the image features into pre-defined categories.

Abstract: A pattern identification apparatus for identifying one of a plurality of classes defined in advance, to which data of a pattern identification target belongs, comprises a read unit adapted to read out, from a storage unit in correspondence with each of the plurality of classes, a projection rule to a hyperplane which approximates a manifold corresponding to the class in a feature space an input unit adapted to input identification target data; a calculation unit adapted to calculate, for each class, a projection result obtained by projecting the input identification target data to the hyperplane which approximates the manifold corresponding to each of the plurality of classes, on the basis of the projection rule; and an identification unit adapted to identify, on the basis of the projection result of each classes calculated by said calculation unit, one of the plurality of classes to which the identification target data belongs.

Abstract: A method for facilitating semantic event classification of a group of image records related to an event. The method using an event detector system for providing: extracting a plurality of visual features from each of the image records; wherein the visual features include segmenting an image record into a number of regions, in which the visual features are extracted; generating a plurality of concept scores for each of the image records using the visual features, wherein each concept score corresponds to a visual concept and each concept score is indicative of a probability that the image record includes the visual concept; generating a feature vector corresponding to the event based on the concept scores of the image records; and supplying the feature vector to an event classifier that identifies at least one semantic event classifier that corresponds to the event.

Abstract: Described is a technology in which face alignment data is obtained by processing an image using a component-based discriminative search algorithm. For each facial component, the search is guided by an associated directional classifier that determines how to move the facial component (if at all) to achieve better alignment relative to its corresponding facial component in the image. Also described is training of the classifiers.

Abstract: An apparatus for efficiently recognizing a part of a body shown in each of plural axial images for one series of axial images obtained by imaging an object to be inspected with a modality. The apparatus includes: apart determining unit for tentatively determining a part of a body shown in each of plural axial images; and a part correcting unit for correcting the part tentatively determined for at least one axial image by the part determining unit, based on information on the plural axial images.

Abstract: A method for measuring multimedia communication quality is disclosed. The multimedia video communication quality may be objectively reflected through the embedment and extraction of digital watermark under a precondition that the quality of the multimedia video data is not obviously affected. In the invention, each frame of the multimedia video data is uniformly divided into blocks of equal size and watermark data is embedded in each of the blocks, so that the watermark may be uniformly distributed. The multimedia video data are divided into groups, and the watermark is embedded in a part of the frames with equal interval between the frames in each group to reduce the effect of the watermark on the data. The watermark information is directly embedded in the spatial domain of the original video data.

Abstract: There is disclosed a label image editing apparatus for editing a label image on a recording media having a data side on which data is recorded, and a label side on which a label image is plotted to visually confirm recording contents on the data side. The label image editing apparatus comprises: a data analyzing section that analyzes data to be recorded on the recording media and/or data to be reproduced from the recording media; and a layout display section that displays, as a layout of the label image, a layout according to an analyzing result by the data analyzing section.

Abstract: Systems and methods for character-by-character alignment of two character sequences (such as OCR output from a scanned document and an electronic version of the same document) using a Hidden Markov Model (HMM) in a hierarchical fashion are disclosed. The method may include aligning two character sequences utilizing multiple hierarchical levels. For each hierarchical level above a final hierarchical level, the aligning may include parsing character subsequences from the two character sequences, performing an alignment of the character subsequences, and designating aligned character subsequences as the anchors, the parsing and performing the alignment being between the anchors generated from an immediately previous hierarchical level if the current hierarchical level is below the first hierarchical level. For the final hierarchical level, the aligning includes performing a character-by-character alignment of characters between anchors generated from the immediately previous hierarchical level.

Abstract: Machine-readable media, methods, apparatus and system for obtaining and processing image features are described. In some embodiments, groups of training features derived from regions of training images may be trained to obtain a plurality of classifiers, each classifier corresponding to each group of training features. The plurality of classifiers may be used to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is. Then, a weight may be discarded from the plurality of weights based upon a certain criterion.

Abstract: Techniques are described for detecting anomalous events using a long-term memory in a video analysis system. The long-term memory may be used to store and retrieve information learned while a video analysis system observes a stream of video frames depicting a given scene. Further, the long-term memory may be configured to detect the occurrence of anomalous events, relative to observations of other events that have occurred in the scene over time. A distance measure may used to determine a distance between an active percept (encoding an observed event depicted in the stream of video frames) and a retrieved percept (encoding a memory of previously observed events in the long-term memory). If the distance exceeds a specified threshold, the long-term memory may publish the occurrence of an anomalous event for review by users of the system.

Abstract: An interest point detection technique is presented. More particularly, for each of possibly multiple image pyramid resolutions, a cornerness image is generated. One or more potential interest point locations are identified in the cornerness image. This involves finding locations associated with a pixel that exhibits a higher corner strength value than pixels in a prescribed-sized surrounding pixel neighborhood. The potential interest point locations are then clustered to identify groups that likely derive from a same 2D structure. Potential interest point locations in one or more of the identified groups are respectively combined to produce a single location that represents the combined group. The representative location of each group having one is then designated as an interest point. An optional location refinement can also be implemented.

Abstract: A long-term memory used to store and retrieve information learned while a video analysis system observes a stream of video frames is disclosed. The long-term memory provides a memory with a capacity that grows in size gracefully, as events are observed over time. Additionally, the long-term memory may encode events, represented by sub-graphs of a neural network. Further, rather than predefining a number of patterns recognized and manipulated by the long-term memory, embodiments of the invention provide a long-term memory where the size of a feature dimension (used to determine the similarity between different observed events) may grow dynamically as necessary, depending on the actual events observed in a sequence of video frames.

Abstract: A method for detecting an image corresponding to at least a monetary banknote is provided. The method includes selecting a detection procedure from a plurality of different detection procedures, and verifying the image according to the selected detection procedure. A method for performing a currency type analysis operation on an image corresponding to at least a monetary banknote is provided. The method includes respectively assigning a plurality of bit values to a plurality of currency types to create a currency type definition table, performing a color analysis operation on the image to generate a color analysis operation result, and determining a currency type of the monetary banknote according to the color analysis operation result and the currency type definition table.

Abstract: The present disclosure related generally to audio and video signal processing. Various arrangements are disclosed. One method recites: (a) obtaining data representing audible portions of audio or representing picture portions of video; (b) using a programmed electronic processor, determining identifying information from the obtained data by computing a frequency transform to produce frequency transform data, and processing the frequency transform data to derive a pattern, and using the pattern as the identifying information for the audio or video; and (c) using the identifying data to facilitate purchase or license of the audio or video. Other arrangements are disclosed as well.

Abstract: In a method and a system for the implementation of multi-layered network object recognition in multi-dimensional space, the structure of a neural recognition network is dynamically generated and adapted to recognize objects. The layers of the network are capable of recognizing key features of the input data by using evaluation rules to establish a hierarchical structure that is independent of data position and orientation, and can adapt varying data densities, geometrical scaling, and faulty or missing data. Adjacent layers of the hierarchy are mutually reinforcing to facilitate the convergence of a solution. Information flow is both bottom-up and top-down during the recognition process providing feedback from higher hierarchical layers to lower layers to cascade the results of higher-level recognition decisions to elements in lower layers.

Abstract: Methods and systems for comparing and organizing color themes and word tag associations. One embodiment comprises a method for determining associated color themes based on an identified color theme by determining the distance between the identified color theme and each color theme of the collection of color themes, wherein each distance includes a color-based distance and the determined subset of associated color themes from the collection is based at least in part on the calculated distances from the identified color theme. Another embodiment comprises a method that allows an application to suggest tags for an identified color theme based on its similarity to color themes and associated tags of the color theme collection. Another embodiment suggests color themes based on an identified tag, and yet another embodiment suggests tags based on an identified tag.

Abstract: Processing data which belong to different classes and label data indicating the classes to which the processing data belong are input (S20). A distance relationship between the processing data is calculated (S22). An interclass separation degree between the classes is set (S23). The distance relationship is updated based on the label data and the interclass separation degree (S24). A data mapping relationship which approximates the updated distance relationship is calculated (S25).

Abstract: Method for at least one of model-based classification and target recognition of an object. The method includes recording an image of an inanimate object, determining a feature that represents a part of the inanimate object, determining at least one condition associated with the feature that indicates an applicability of the feature based on at least one of geometry of the object, distance of the object from a camera, illumination conditions, contrast, speed of the object, height of the object, and relative position of the object to a camera, and carrying out the at least one of classification and target recognition of the object by recording the feature when the at least one condition indicates the applicability of the feature. At least one of object classification and target recognition is carried out for a feature of the object that is visible and recordable according to the position of the object.

Abstract: Disclosed are an apparatus and a method for determining a multi-view specific object. The apparatus comprises an input device for inputting image data; and cascade classifiers formed of stage classifiers corresponding to a same detection angle, the stage classifiers corresponding to different features. Each cascade classifier is for calculating a degree of confidence of the image data of a specific object corresponding to the detection angle based on the aspect of the corresponding feature, and determining whether the image data belongs to the specific object based on the degree of confidence. A self-adaptive posture prediction device is disposed between two stage classifiers in each cascade classifier, and is used to determine whether the image data enters the cascade classifiers corresponding to the detection angles and located after the self-adaptive posture prediction device.

Abstract: Apparatus, methods, systems and devices for identifying and tracking a human hand in a video. The steps for identifying the hand includes detecting parallel lines (bars), detecting curved fingertips (curves), grouping the detected parallel lines and curved fingertips according to a parallel line decision tree and a curved fingertip decision tree, respectively, and merging the parallel line group and the curved fingertip group to identify candidate hands.

Abstract: Provided are systems, apparatuses, methods and techniques for recognizing a pattern within input data, such as a specific object within an image frame. Prior to normalizing the input data, normalization-independent classification is performed, thereby rejecting a number of potential windows early on and resulting in a corresponding reduction in required processing.