Abstract: Systems, methods, and machine-readable media for generating a classifier configured to label segments of an image, are discussed. According to one aspect, the system may include a training module, a labeling module, and an update module. The training module may be configured to train a first sub-classifier based on photographic data for a set of pre-labeled image segments and a second sub-classifier based on 3-dimensional point data for the set of pre-labeled image segments. The labeling module may be configured to generate a labeling solution comprising a plurality of associations between an image segment from the set of unlabeled image segments and a label. The update module may be configured to update the set of pre-labeled image segments based on the labeling solution. The training module may also be configured to train the first sub-classifier and the second sub-classifier based on the updated set of pre-labeled image segments.

Abstract: Methods and systems for generating unbiased wafer defect samples are provided. One method includes selecting the defects detected by each of multiple scans performed on a wafer that have the most diversity in one or more defect attributes such that a diverse set of defects are selected across each scan. In addition, the method may include selecting the defects such that any defect that is selected and is common to two or more of the scans is not selected twice and any defects that are selected are diverse with respect to the common, selected defect. Furthermore, no sampling, binning, or classifying of the defects may be performed prior to selection of the defects such that the sampled defects are unbiased by any sampling, binning, or classifying method.

Abstract: An apparatus and method for filtering depth information received from a capture device. Depth information is filtered by using confidence information provided with the depth information based an adaptively created, optimal spatial filter on a per pixel basis. Input data including depth information is received on a scene. The depth information comprises a plurality of pixels, each pixel including a depth value and a confidence value. A confidence weight normalized filter for each pixel in the depth information is generated. The weight normalized filter is combined with the input data to provide filtered data to an application.

Abstract: A method for determining a matching score between a first set of n1 feature points, and a second set of n2 feature points includes producing a triple-wise affinity tensor, determining a leading eigenvector of the triple-wise affinity tensor, iteratively producing a binary optimal assignment vector and determining a matching score between the first set of feature points and the second set of feature points. The triple-wise affinity tensor includes the affinity score of assignments of triplets of feature points of the first set of feature points and triplets of feature points of the second set of feature points. The procedure of iteratively producing a binary optimal assignment vector is performed by discretization of the leading eigenvector. The procedure of determining a matching score is performed according to the triple-wise affinity tensor and according to the optimal assignment vector.

Abstract: A method and device for acquiring latent fingerprint orientation. The method includes: extracting a valid area in a latent fingerprint image; acquiring an orientation field of the valid area and a confidence of coordinate points in the valid area; dividing the orientation field into a first area with a high confidence and a second area with a low confidence; collecting control points in the first area, performing triangulation on control points in the first area and determining differences between control points to determine a set of control points; manually marking control points in the second area, adding control points marked to the set, performing triangulation on control points in the set and determining whether a new manually marked control point is to be added in the set according to a difference determination; and establishing a model for control points in the set to obtain an orientation field of the image.

Abstract: Enhanced accuracy finger position and motion sensors devices, algorithms, and methods are disclosed that can be used in a variety of different applications. The sensors can be used in conjunction with partial fingerprint imagers to produce improved fingerprint scanners. Such improved scanners can use image analysis techniques, such as interpolation between partial fingerprint images to correct for missing data, or discarding redundant partial fingerprint image data, to produce adequate fingerprint images even when the finger has not been applied to the sensor using an optimum technique.

Abstract: The invention provides methods for automated classification of a medical diagnostic image of a lung according to its deduced probability of relating to a lung of a patient who is suffering from a diffuse parenchymal lung disease such as chronic obstructive pulmonary disease (COPD), cystic fibrosis, or severe asthma, or to a class of patients characterized by the severity of such a condition, or to a class of patients characterized by a prognostic likelihood of developing such a condition or severity of condition.

Abstract: The described implementations relate to computer vision. In one case image data is received. Resource constraints associated with processing the image data are evaluated. Further, a visual recognition and detection processing strategy is selected for the image data based at least in part on the evaluated resource constraints.

Abstract: A biological sample such as a tissue section is stained with one or more quantum dots and possibly other fluorophores (total number of fluorophores N). A camera coupled to a microscope generates an image of the specimen at a plurality of different wavelengths within the emission spectral band of the N fluorophores. An analysis module calculates coefficients C1 . . . CN at each pixel which are related to the concentration of each of the individual fluorophores. Morphological processing instructions find biological structures. A display module displays quantitative analysis results to the user. The quantitative analysis display includes histograms of the biological structures, scatter plots of fluorophore concentrations, statistical data, spectral data, image data and still others.

Abstract: An image region extracting unit binarizes image data inputted from an image inputting unit and labels the binarized image data based upon a feature value of a pixel to be segmentized into a plurality of image regions. A contour connection determining unit 104 determines whether or not connection points connecting contour lines of the respective image regions extracted by a contour extracting unit each other exist. A contour modifying unit, when it is determined that the connection points exist by the contour connection determining unit, connects the contour lines each other at the connection points to modify the contour line.

Abstract: Precise boundary segment intersection methods and systems for boundary representation are described. For example, in one embodiment, a method includes computing a plurality of intersections of a plurality of segments, computing an uncertainty associated with each intersection of the plurality of intersections, and determining whether to aggregate at least one of the plurality of intersections with another of the plurality of intersections based on one or more of the determined uncertainties. In another embodiment, the method may further include determining whether to consolidate at least one non-aggregated intersection with at least one other intersection based on a topological consistency requirement.

Abstract: A method of detecting streaks in printed documents comprises, with an imaging device, capturing an image of a printed document, transmitting the image data representing the printed document to a computing device, with the computing device, dividing the image data into a number of rectangles horizontally across the image data so that the rectangles overlap in the horizontal direction, and with the computing device, determining if a rectangle comprises a streak. A system for detecting and evaluating streaks in printed document comprises a printing device that prints a document based on image data received from a computing device, an imaging device that captures an image of the printed document, and a streak detector that detects streaks within the captured image of the printed document.

Abstract: An apparatus and method for human activity and facial expression modeling and recognition are based on feature extraction techniques from time sequential images. The human activity modeling includes determining principal components of depth and/or binary shape images of human activities extracted from video clips. Independent Component Analysis (ICA) representations are determined based on the principal components. Features are determined through Linear Discriminant Analysis (LDA) based on the ICA representations. A codebook is determined using vector quantization. Observation symbol sequences in the video clips are determined. And human activities are learned using the Hidden Markov Model (HMM) based on status transition and an observation matrix.

Abstract: A method of apprehending a criminal, including scanning fingerprint images from a contact area of at least one item at a point-of-sale (POS) location and/or point-of-entry (POE) location and transmitting the fingerprint images to a local or remote electronic device to identify the criminal is presented. A fingerprint recognition and collection device is also presented including a fingerprint scanner for scanning fingerprint images from a contact area of at least one item; a transmitting unit for transmitting the fingerprint images; a receiving unit for receiving the fingerprint images; and an analyzing unit for analyzing the fingerprint images to identify persons of interest. The fingerprint scanner operates concurrently with a price scanner and/or ticket scanner. The fingerprint scanner is fixedly secured at the POS location and/or the POE location and operates concurrently with the price scanner and/or ticket scanner.

Abstract: A method of managing visiting guests by face recognition includes: extract the face image of the visiting guest; determine if the face image of the guest is stored in the database; display the face image data comprising the face image, guest information and visiting information if the face image is stored or enroll the new face image data and then display the face image data just registered; set up at least one authorized item; execute the visiting goal confined by the authorized items; monitor the status of the visiting guest; wait until the guest leaves; and finally record and update log page and face image data of the visiting guest in the database. Therefore, the method can reduce security guard workload, assist security guards or automatically and solely execute the function of guard entrance without any careless mistakes, and update visiting guest information in real time.

Abstract: Embodiments of the present invention relate to a method and a system for personalized advertisement push based on user interest learning. The method may include: obtaining multiple user interest models through multitask sorting learning; extracting an object of interest in a video according to the user interest models; and extracting multiple visual features of the object of interest, and according to the visual features, retrieving related advertising information in an advertisement database. Through the method and the system provided in embodiments of the present invention, a push advertisement may be closely relevant to the content of the video, thereby meeting personalized requirements of a user to a certain extent and achieving personalized advertisement push.

Abstract: A wordspotting system and method are disclosed for processing candidate word images extracted from handwritten documents. In response to a user inputting a selected query string, such as a word to be searched in one or more of the handwritten documents, the system automatically generates at least one computer-generated image based on the query string in a selected font or fonts. A model is trained on the computer-generated image(s) and is thereafter used in the scoring the candidate handwritten word images. The candidate or candidates with the highest scores and/or documents containing them can be presented to the user, tagged, or otherwise processed differently from other candidate word images/documents.

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: An optics detection system is disclosed. The optics detection system includes a sensor module and a processor. The sensor module is configured to illuminate a field of regard with a plurality of light pulses and to capture reflections of the plurality of light pulses in a plurality of frames, respectively. The processor is configured to process the plurality of frames to locate and identify optics within the field of regard using a plurality of discriminators.

Abstract: A non-contact passive ranging system wherein a first imager on a platform is focused on a first object and a second imager on the platform is also focused on the first object. The optical path from the first object to the first imager is configured to be shorter than the optical path from the object to the second imager. Processing circuitry is responsive to an output of the first imager and an output of the second imager as relative motion is provided between the platform and the first object and is configured to calculate the distance from the platform to the object.