Abstract: Automated systems, methods and tools that automatically extract and select portions of an image to automatically generate a premium finish mask specific to the image which require little or no human intervention are presented. Graphical user interface tools allowing a user to provide an image and to indicate regions of the image for application of premium finish are also presented.

Abstract: An identification apparatus performs classification using a plurality of classifiers, and calculates the reliability of its classification result. A data obtaining unit obtains input data. A feature quantity obtaining unit obtains a feature quantity corresponding to the input data. A plurality of classifiers receive input of the feature quantity and perform classification based on the input feature quantity. An identification unit inputs the feature quantity into each of the classifiers, and generates a single second classification result based on a plurality of classification results obtained from the classifiers. A reliability generation unit generates a reliability of the second classification result based on variations across the plurality of classification results.

Abstract: An image processing device includes a feature point detecting section which detects a feature point from an image and which includes a ternarizing section and a determining section. The ternarizing section sets each pixel of the image as a target pixel and inputs pixel values of the target pixel and surrounding pixels. The ternarizing section changes each of the pixel values of the surrounding pixels to one of three signal values which represent that brightness of a surrounding pixel is brighter or darker than or similar to brightness of the target pixel. When the number of consecutive surrounding pixels having the same signal value is equal to or greater than a threshold, the determining section determines the target pixel as the feature point.

Abstract: A Robotic control system has a wand, which emits multiple narrow beams of light, which fall on a light sensor array, or with a camera, a surface, defining the wand's changing position and attitude which a computer uses to direct relative motion of robotic tools or remote processes, such as those that are controlled by a mouse, but in three dimensions and motion compensation means and means for reducing latency.

Abstract: A system for transforming an image of a fingerprint, comprises a mobile device, comprising: a first communication interface, a camera configured to capture at least one image of at least one fingerprint, and a mobile device processor configured to execute instructions, the instructions configured to cause the mobile device processor to receive the image from the camera and transmit them to an image processing system via the first communication interface; and an image processing system, comprising: a second communication interface configured to receive the image, and an image processor configured to execute instructions, the instructions configured to cause the image processor receive the image form the second communication interface, and: render the image into a high contrast image, establish focus and image resolution for the image, perform noise reduction on the image, and perform distortion elimination on the image.

Abstract: Techniques for assigning context to a digitally captured image, and for manipulating recognized data fields within such image. In an exemplary embodiment, a context of an image may be assigned based on, e.g., user input or pattern recognition. Based on the assigned context, recognized data fields within the image may be manipulated according to context-specific processing. In an aspect, processing specific to a sales receipt context may automatically manipulate certain data, e.g., calculate updated sales tax and subtotals based on user-designated fields, and display the automatically calculated data in an output receipt. Fields not designated by the user may be selectively concealed in the output receipt for privacy. Further aspects disclose processing techniques specific to other contexts such as restaurant menu, store shelf, and fillable form contexts.

Abstract: Disclosed herein is a method for providing image contents. The method for providing image contents includes: dividing the image contents into a plurality of scenes, each scene including a plurality of shots; classifying image frames for each scene depending on each of a plurality of characters appearing in the image contents; receiving a user input for selecting any one of the plurality of characters; and displaying a scene corresponding to the character selected depending on the user input.

Abstract: An apparatus for detecting a feature in an image includes an image input section for receiving at least part of the image in the form of image data having a plurality of pixels, the plurality of pixels comprising a plurality of non-border pixels, a feature detection module adapted to attribute a feature probability value to each of the pixels of the image data, and an extremum determination module for determining at least one local extremum among the feature probability values, wherein the extremum determination module is adapted to output, for each of the plurality of pixels, a final indication if the feature probability value of the pixel in question is a local extremum.

Abstract: An electronic device and fingerprint recognition method are provided. The electronic device includes a housing, a lens protection cover, a processing unit, an image capturing unit, a light-source sensor and a visible light source module. The lens protection cover is installed on the housing. The light-source sensor is configured to detect external light which comes into the electronic device through the lens protection cover. The processing unit determines whether the detected external light is less than a threshold. The visible light source module is configured to generate visible light. When the detected external light is less than the threshold, the processing unit activates the visible light source module to generate the visible light, and activates the image capturing unit to capture a fingerprint image of a user's finger which is put on the lens protection cover for fingerprint recognition.

Abstract: A method for generating a sharp image based on a blurry image is provided. The method includes acquiring pixel values of pixels in the blurry image, and a convolution kernel of the blurry image, determining a deconvolution kernel of the blurry image based on a preset image gradient operator and the convolution kernel, determining pixel values of pixels in the sharp image based on the deconvolution kernel and the pixel values of the pixels in the blurry image, and generating the sharp image based on the pixel values of the pixels in the sharp image. A deconvolution kernel in the embodiments of the present disclosure is determined based on a gradient operator and a convolution kernel, in other words, the deconvolution kernel introduces the gradient operator as a regularization constraint, which prevents noise from affecting an image recovery process, and improves the quality of a recovered sharp image.

Abstract: A machine-learning engine is disclosed that is configured to recognize and learn behaviors, as well as to identify and distinguish between normal and abnormal behavior within a scene, by analyzing movements and/or activities (or absence of such) over time. The machine-learning engine may be configured to evaluate a sequence of primitive events and associated kinematic data generated for an object depicted in a sequence of video frames and a related vector representation. The vector representation is generated from a primitive event symbol stream and a phase space symbol stream, and the streams describe actions of the objects depicted in the sequence of video frames.

Abstract: Techniques related to depth image enhancement for hardware generated depth images are discussed. Such techniques may include determining a depth image enhancement indicator for a depth image based on the depth image and a depth image model and generating pixel depth values for missing pixel depth values of the depth image when the depth image enhancement indicator indicates enhancement. The pixel depth values may be generated based on search windows that extend from a position greater than a predetermined disparity position to a maximum disparity position.

Abstract: Optical image data is acquired by irradiating a pattern with light emitted from a light source. A threshold value is specified by internally dividing a minimum value and a maximum value of a signal amount of reference image data by a division ratio. A position corresponding to a signal amount of a threshold value is determined as an edge of a pattern of the reference image data. A position of a signal amount equal to the threshold value is determined as an edge of the pattern of the optical image data. A line width error is obtained as a difference between a first line width of the optical image data and a second line width of the reference image data. A new threshold value is specified in the case of fluctuation of a light quantity of the light source or decrease of a contrast value of the optical image data.

Abstract: Systems and methods for training a machine vision system create geometric models. The disclosed methods can extract one or more corresponding features and one or more differentiating features from different sets of training images. The one or more differentiating features can be used to differentiate between the different work pieces. The disclosed methods can generate an alignment model using the corresponding features and a classification model using the one or more differentiating features.

Abstract: Disclosed is a sampling method for estimating a homography matrix. This sampling method, for estimating a homography matrix that represents conversion correlations between pluralities of images by means of Constraint Satisfaction-Random Sample Consensus (CS-RANSAC), includes the steps of: sampling to divide an input image into a form of grids (N by N), select features, which are used for calculating a homography matrix, from features, which are abstracted from the input image, by means of a random sampling, and inspect whether the features selected by the random sampling satisfy predefined constraints; and executing model estimation to calculate the homography matrix from the features only if the selected features satisfy the constraints.

Abstract: A computer-implemented system and method for generating and placing cluster groups is provided. A set of clusters each having one or more documents is maintained. A portion of the clusters is assigned to two or more spines. The cluster spines are ordered by cluster length. Those cluster spines that are unique are identified in decreasing cluster length order and placed. One or more of the remaining spines are placed in relation to at least one of the placed unique spines. At least one of the remaining clusters is placed in relation to one or more of the unique placed spines. Groups of clusters are displayed. Each group includes two or more of one such unique placed spine, the remaining spines placed in relation to that placed unique spine; and the remaining clusters placed in relation to the unique placed spine.

Abstract: In one embodiment, a method for detecting faces in video image frames is implemented on a computing device including: comparing current image frames to previously processed image frames to determine similarity, if a current image frame and a previously processed image frame are dissimilar, comparing a location within the current image frame for at least one detected facial image to a location within an associated image frame for at least one most recently stored facial image stored in a most recently used (MRU) cache, if the compared locations are dissimilar, comparing the at least one detected facial image to the at least one most recently stored facial image stored in the MRU cache to determine similarity, and storing the at least one detected facial image in the MRU cache if the at least one detected facial image and the at least one most recently stored facial image are not dissimilar.

Abstract: A handwriting recognition system converts word images on documents, such as document images of historical records, into computer searchable text. Word images (snippets) on the document are located, and have multiple word features identified. For each word image, a word feature vector is created representing multiple word features. Based on the similarity of word features (e.g., the distance between feature vectors), similar words are grouped together in clusters, and a centroid that has features most representative of words in the cluster is selected. A digitized text word is selected for each cluster based on review of a centroid in the cluster, and is assigned to all words in that cluster and is used as computer searchable text for those word images where they appear in documents. An analyst may review clusters to permit refinement of the parameters used for grouping words in clusters, including the adjustment of weights and other factors used for determining the distance between feature vectors.

Abstract: A system and method are provided for reconstructing images from limited or incomplete data, such as few view data or limited angle data or truncated data generated from divergent beams. The method and apparatus may iteratively constrain the variation of an estimated image in order to reconstruct the image. To reconstruct an image, a first estimated image may be generated. Estimated data may be generated from the first estimated image, and compared with the actual data. The comparison of the estimated data with the actual data may include determining a difference between the estimated and actual data. The comparison may then be used to generate a new estimated image. For example, the first estimated image may be combined with an image generated from the difference data to generate a new estimated image. To generate the image for the next iteration, the variation of the new estimated image may be constrained.

Abstract: An optical module includes a housing with a lighting mechanism, an aperture formed therein, and a window that frames a transparent surface adapted to contact a skin surface of a person, for example, a fingertip. The lighting mechanism provides light to illuminate the skin surface placed upon the transparent surface. A prism has a first side facing the lighting mechanism, a second side at the window, and a third side through which a dactylographic image exits the prism. One or more light reflecting surfaces are disposed within the housing to reflect the dactylographic image towards the housing aperture. The optical module is coupled to a mobile device having a camera, with the aperture of the housing aligning with a lens of the camera. The camera acquires the dactylographic image, and the mobile device adjusts this dactylographic image to produce a dactylographic image suitable for biometric matching.