Abstract: Methods, systems, and computer readable media for fingerprint capture and verification are described. A camera is utilized to capture a set of images at a first resolution. A focal length for at least one object in the plurality of images is verified to be greater than a minimum focus threshold. The images are converted to corresponding binary images. A finger is identified in the one or more binary images. A still image of the finger is captured at a second and higher resolution. Image processing filters are applied to the still image to generate a fingerprint image which is transmitted to an application server for verification.

Abstract: Techniques are disclosed for image matting. In particular, embodiments decompose the matting problem of estimating foreground opacity into the targeted subproblems of estimating a background using a first trained neural network, estimating a foreground using a second neural network and the estimated background as one of the inputs into the second neural network, and estimating an alpha matte using a third neural network and the estimated background and foreground as two of the inputs into the third neural network. Such a decomposition is in contrast to traditional sampling-based matting approaches that estimated foreground and background color pairs together directly for each pixel. By decomposing the matting problem into subproblems that are easier for a neural network to learn compared to traditional data-driven techniques for image matting, embodiments disclosed herein can produce better opacity estimates than such data-driven techniques as well as sampling-based and affinity-based matting approaches.

Abstract: Methods and systems of inserting image files into a container are described herein. One method includes overseeing a submission of images to containers, retrieving a job of inserting images into the container, and inserting images into the container by, for each of the plurality of images: (i) reading and analyzing source data including the image; (ii) opening the container, which includes records, where the image will be stored; (iii) determining whether the source data is valid to be entered into a record; (iv) responsive to a determination that the source of data is valid, preparing metadata; (v) writing the record into the container; and (vi) marking a status flag/mark of the record as being updated while writing the record into the container.

Abstract: Image blurring methods and apparatuses, storage media, and electronic devices can include: obtaining a main image and a secondary image obtained by photographing the same object with a dual-lens camera; obtaining depth data and depth confidence degree data according to the main image and the secondary image, the depth data indicating depth values of corresponding pixel points in the main image and the secondary image, and the depth confidence degree data indicating confidence degrees of the depth values in the depth data; correcting at least one depth value in the depth data according to the depth confidence degree data; and blurring the main image according to corrected depth data.

Abstract: An information processing apparatus includes a storage unit that previously stores a classification rule corresponding to an attribute of an applicant, an acquisition unit that acquires content of an electronic document from an image of the document, and a setting unit that reads the classification rule corresponding to the attribute of the applicant of the document from the storage unit, and sets a classification item of the document based on the content of the document, which is acquired by the acquisition unit, and the classification rule which is read from the storage unit.

Abstract: Methods, systems and apparatuses of feature extraction and object detection are provided. In the method of feature extraction, a plurality of image channels are generated from each of training images; intra-channel features are extracted from the plurality of image channels for each of training images, wherein the intra-channel features include features independently extracted from a single image channel; cross-channel features are extracted from the plurality of image channels for at least one of the training images, wherein the cross-channel features include features extracted from at least two image channels. The intra-channel features and the cross-channel features form a set of features for feature selection and classifier training.

Abstract: A method and device for determining the color homogeneity of hair is provided. In accordance with various exemplary embodiments, a method comprises providing a digital image in which hair is depicted and which is parameterized in a color space which is defined by employing a parameter set in which one of the parameters is a hue, defining at least one hair analysis region in the digital image, determining a hue distribution in at least one hair analysis region, and determining at least one color homogeneity value on the basis of the determined hue distribution.

Abstract: A method and a system of capturing an image of a card having a magnetic stripe is provided. The method includes obtaining a first image by an imaging device of the card, obtaining a plurality of images of the card via color delta analysis, and obtaining a third image of the card by comparing the first and the plurality of images.

Abstract: A device for image processing includes an optical receiver configured to receive a reflection of a coded pattern from an object to generate an image, and processing circuitry. The processing circuitry is configured to determine an estimated position of zero order light in the image, determine a spatial region of the coded pattern that corresponds to a position of the zero order light in the coded pattern, map the spatial region to the estimated position of the zero order light in the image to generate a corrected image, and generate a depth map for the coded pattern based on the corrected image.

Abstract: The presence of an object in a query image may be searched for using a feature of the query image. A local descriptor value for the feature is first determined. Previously stored object data for various 3D objects is then retrieved, with each set of stored object data including a probability distribution that describes a feature of a 3D object as obtained from images at different camera poses. A probability value is then determined using the local descriptor value for the feature in the query image and the available distribution data from the stored object data. The distribution data is then ranked based on the resulting probability values and the highest ranked distribution data is associated with the feature in the query image. Output data may also be generated to identify the 3D object in the query image based on an object identifier of the highest ranked distribution data.

Abstract: Dynamic image content is generated based on various combinations of image elements associated with an input image unit. In this regard, an input image unit is selected and input into a dynamic content generation engine. The input image unit includes a number of image elements. Different combinations of image elements in the input image are added and/or removed to generate candidate image units. Different colors may be also be assigned to image elements based on a color palette. In this way, permutatively different candidate image units are automatically generated with different combination of elements from the input image unit and possibly different colors. Generation of candidate image units can be based on the application of a combination formula onto the image elements associated with the input image unit. The candidate image units are then displayed for selection and further modification.

Abstract: The present invention provide a full-servo label inspection machine and a control method thereof; the machine includes a frame, the frame is provided with an unrolling mechanism, a rolling mechanism, a visual inspection mechanism, a label removal mechanism, and a label replacement mechanism. In the label removal mechanism, when the paper pressing roller is at the label removal position, it presses the material to be inspected downwards to form an angle of less than 180° with the rear end of a label removal plate. After a defective label on the material to be inspected passes the label removal plate, a front end of the defective label curls up and eventually sticks onto the label removal paper. The present invention allows fully automation of the removal and replacement of labels, achieving high inspection efficiency and accuracy.

Abstract: The present invention has disclosed a multifunctional automatic quality inspection machine and a control method thereof. The machine comprises a frame, an unrolling device provided at the frame, an inspection mechanism, a button operating platform, a slitting mechanism, a main controller, a display device, and a rolling device; a paper movement route is provided between the unrolling device and the rolling device; the inspection mechanism and the display device are respectively electrically connected to the main controller; the rolling device includes a first rolling airshaft, a second rolling airshaft, and a rolling drive device; the first rolling airshaft and the second rolling airshaft are provided on the rolling drive device; the first rolling airshaft and the second rolling airshaft swap their positions via the rolling drive device. The machine has a number of functions including automatic unrolling, high inspection precision, discharging waste materials, splitting and quick rolling.

Abstract: A method of training a captioning model used to perform automatic video captioning of an input video, including initializing, by at least one processor, a plurality of long short-term memory (LSTM) units included in the captioning model using cross-entropy loss; training, by the at least one processor, the LSTM units using reinforcement learning; training, by the at least one processor, the LSTM units and a plurality of convolutional neural networks (CNNs) included in the captioning model using multitask training; and generating, by the at least one processor, a video caption corresponding to the input video using the captioning model.

Abstract: A method of segmenting an image of a scene captured using one of a plurality of cameras in a network. A mask of an image of a scene captured by a first one of said cameras is received. A set of pixels in the mask likely to be in a foreground of an image captured by a second one of said cameras is determined based on the received mask, calibration information, and a geometry of the scene. A set of background pixels for the second camera is generated based on the determined set of pixels. The generated set of background pixels is transmitted to the second camera. The image of the scene captured by the second camera is segmented using the transmitted background pixels.

Abstract: A shape discriminating device includes an image acquisition means configured to acquire a target image being an image where pants are shown, a diverging position acquisition means configured to acquire, as a diverging position, a position where the pants diverge into two leg parts or a position which can be regarded as a position where the pants diverge into two leg parts, a leg region extraction means configured to extract a leg region from a region showing the pants in the target image, a leg shape acquisition means configured to acquire leg shape information concerning variation in lateral width of a leg part of the pants based on the diverging position and the leg region, a specifying means configured to specify a shape of the pants based on the leg shape information, and an output means configured to output information concerning the specified shape of the pants.

Abstract: Methods and apparatus for using features of images representing content items to improve the presentation of the content items are disclosed. In one embodiment, a plurality of digital images are obtained, where each of the images represents a corresponding one of a plurality of content items. Image features of each of the digital images are determined. Additional features including at least one of user features pertaining to a user of a client device or contextual features pertaining to the client device are ascertained. At least a portion of the content items are provided via a network to the client device using features that include or are derived from both the image features of each of the plurality of digital images and the additional features.

Abstract: A component supply device that images stage on which multiple leaded components are supported in a scattered state, and holds a component supported on the stage using a component holding tool based on the image data, wherein multiple stages of different colors are prepared, and a stage from among the multiple stages a stage with a color different to a component that is planned to be supplied is removably attached to component support member. By this, for example, in a case in which a component planned to be supplied is white, by attaching a black stage to the component support section, due to the contrast between the background and the target object, it is possible to clearly recognize the white component supported on the black stage.

Abstract: An information processing apparatus includes an obtaining unit that obtains image data which is data obtained by digitizing a document including plural areas in one document, a setting unit that sets a determination standard for each of the areas based on a degree of necessity of a confirming operation for a text or a picture included in the area of the image data, and a determination unit that determines a method of the confirming operation for each of the areas based on the determination standard.

Abstract: Low power input object settlement detection systems and methods for operating a capacitive sensor having a plurality, M, of transmitter electrodes and a plurality, N, of receiver electrodes, wherein N and M are integer values. A plurality of input object settlement scans are captured, when a presence of an input object is detected, or in response to a presence of an input object being detected, wherein capturing each input object settlement scan includes driving all or a portion of the plurality, M, of transmitter electrodes simultaneously and detecting receiver signals from at least a subset of the plurality, N, of receiver electrodes simultaneously. When a difference between subsequent input object settlement scans is below a threshold value, a full input object image is acquired.