Abstract: A method of performing an iterative bundle adjustment for an imaging device is described. The method comprising implementing a plurality of functions in performing a bundle adjustment. Predetermined functions of the plurality of functions may be started using a processor for a second iteration in parallel with a first iteration of the plurality of functions. A result of the predetermined functions started during the first iteration may be used in a second iteration. An output of the bundle adjustment may then be generated for successive iterations.

Abstract: There is provided an information processing method and an information processing device to provide positional information with higher accuracy in accordance with situations in the real world. The information processing method includes: generating, by a processor, a feature point list in which three-dimensional coordinates of a feature point are associated with a local feature amount of the feature point, the feature point having been detected from observation information collected around a unit area. The information processing device includes: a computation section configured to extract a feature point and a local feature amount related to the feature point from acquired image information; and a communication section configured to acquire a feature point list on a basis of collected observation information.

Abstract: A system can visually track which items in a store are selected for purchase by a shopper. The system can form a virtual shopping cart by analyzing multiple images, over time, to determine which purchasable items are located with the shopper, such as in a physical shopping cart, in a basket, or held by the shopper. By analyzing multiple images, over time, the system can account for items misidentified in one or more images, or fully or partially obscured in one or more images as the shopper traverses the store. Alternatively, the system can form a virtual shopping cart by analyzing instances in which a purchasable item is removed from a shelf or placed on a shelf. Items removed from, but not returned to, a shelf can be considered to be selected for purchase. The system can include a frictionless checkout that charges the shopper for the selected items.

Abstract: The present disclosure relates to simulating the capture of images. In some embodiments, a document and a camera are simulated using a three-dimensional modeling engine. In certain embodiments, a plurality of images are captured of the simulated document from a perspective of the simulated camera, each of the plurality of images being captured under a different set of simulated circumstances within the three-dimensional modeling engine. In some embodiments, a model is trained based at least on the plurality of images which determines at least a first technique for adjusting a set of parameters in a separate image to prepare the separate image for optical character recognition (OCR).

Abstract: An electronic device for image correction and a method therefore are provided. The electronic device includes a camera, a memory, a communication circuit, and at least one processor. The at least one processor is configured to obtain a raw image of an external object using the camera, determine whether correction of the raw image is necessary, when the correction of the raw image is determined not to be necessary, store an image corresponding to the raw image in the memory, and when the correction of the raw image is determined to be necessary, transmit the raw image to an external electronic device via the communication circuit so as to enable the external electronic device to generate a corrected image of the raw image, and store the corrected image in the memory.

Abstract: The invention concerns a method of optical, inspection of an electronic circuit (Card) including the acquisition of images of the electronic circuit by image sensors (C), the use of the images to determine the offset between the position of the electronic circuit (Card) and an inspection position, and the use of said images in at least another step of the method.

Abstract: A system for detecting synthetic videos may include a server, a plurality of weak classifiers, and a strong classifier. The server may be configured to receive a prediction result from each of a plurality of weak classifiers; and send the prediction results from each of the plurality of weak classifiers to a strong classifier. The weak classifiers may be trained on real videos and known synthetic videos to analyze a distinct characteristic of a video file; detect irregularities of the distinct characteristic; generate a prediction result associated with the distinct characteristic, the prediction result being a prediction on whether the video file is synthetic; and output the prediction result to the server. The strong classifier may be trained to receive the prediction results of the plurality of weak classifiers from the server; analyze the prediction results; and determine if the video file is synthetic based on the prediction results.

Abstract: The present invention discloses an image processing apparatus and method and storage medium storing instructions. The image processing apparatus comprises a unit for acquiring a current video frame and a scene model based on an input video, a unit for updating visual element models in said scene model, which match with the visual elements in said current video frame, based on the visual elements in said current video frame, and a unit for updating visual element models in said scene model, which do not match with the visual elements in said current video frame, based on the visual elements in said current video frame and in the previous video frames thereof. According to the present disclosure, the whole scene model can vary immediately following the variation in the environment illumination.

Abstract: Embodiments of the present systems and methods may provide techniques that provide automatic, reliable performance of the point cloud cleanup task. Embodiments may provide the capability to progressively learn the object-background segmentation from tracking sessions in augmented reality (AR) applications. The 3D point cloud may be used for tracking in AR applications, by matching points from the cloud to regions in the live video. A 3D point that has many matches is more likely to be part of the object, while a 3D point that rarely has matches is more likely to be part of the background. Advantages of this approach include that no manual work is needed to do the segmentation, and the results may be constantly improved over time, as the object is being tracked in multiple environments.

Abstract: A chest x-ray differential diagnosis system is operable to generate abnormality pattern data is generated for each of a received plurality of chest x-rays by identifying at least one pattern in each chest x-ray corresponding to an abnormality by utilizing a computer vision model that is trained on a plurality of training chest x-rays. Differential diagnosis data is generated for each chest x-ray based on the abnormality pattern data. Filtering parameters are received from a client device, and a filtered chest x-ray queue that includes a subset of chest x-rays is selected based on the filtering parameters and the differential diagnosis data is generated for transmission to the client device for display. Differential diagnosis data corresponding a chest x-ray indicated in chest x-ray selection data received from the client device is transmitted to the client device for display via the display device in conjunction with the chest x-ray.

Abstract: A multi-pose face feature point detection method based on cascade regression comprises: extracting a pose index features and establishing corresponding optimal weak regressors; performing corresponding initialization according to different face pose orientations; using an SIFT feature of a face image as an input feature for face orientation estimation; acquiring an orientation of an input face image according to a random forest face orientation decision tree; using a feature point mean value of a face training sample under the orientation as an initial value of the input face image feature point; and extracting the pose index feature of the face image and inputting the pose index feature into the optimal weak regressor to acquire a distribution residual to update the current feature point distribution, and complete the face feature point detection.

Abstract: An apparatus includes a sensor, a navigation circuit and a control circuit. The sensor may be configured to generate surrounding road information of a road. The road may have a plurality of available lanes. The navigation circuit may be configured to determine a current position of the apparatus on the road. The control circuit may be configured to (i) access map data that identifies a number of the available lanes in the road proximate the position, (ii) determine a current lane among the available lanes that the apparatus is within based on all of the position, the surrounding road information and the map data and (iii) generate feedback data based on both the position and the current lane. The navigation device may be further configured to adjust the current position to a center of the current lane in response to the feedback data.

Abstract: A technique and system for counting the number of repetitions of approximately the same action in an input video sequence using 3D convolutional neural networks is disclosed. The proposed system runs online and not on the complete video. It analyzes sequentially blocks of 20 non-consecutive frames. The cycle length within each block is evaluated using a deep network architecture and the information is then integrated over time. A unique property of the disclosed method is that it is shown to successfully train on entirely synthetic data, created by synthesizing moving random patches. It therefore effectively exploits the high generalization capability of deep neural networks. Coupled with a region of interest detection mechanism and a suitable mechanism to identify the time scale of the video, the system is robust enough to handle real world videos collected from youtube and elsewhere, as well as non-video signals such as sensor data revealing repetitious physical movement.

Abstract: [Problem] To provide a highly accurate cell measurement method. [Solution] A cell measurement method comprising: a step of staining target cells with a dye; an image obtaining step for obtaining an image of the target cells; a discrimination step for discriminating the target cells from contaminating cells by applying multi-stage binarization processing to the image; a step of eliminating noises due to the contaminating cells from the image based on the result of the discrimination step; and a step of evaluating an amount of target cells by integrating an index value of cell amount in the image from which the contaminating cells have been eliminated.

Abstract: A super-resolution lattice light field microscopic imaging system includes: a microscope configured to magnify a sample and image the sample onto a first image plane of the microscope; a first relay lens configured to match a numerical aperture of an objective with that of a microlens array; a 2D scanning galvo configured to rotate an angle of a light path in the frequency domain plane; an illuminating system configured to provide uniform illumination on the microlens array to generate SIM pattern illumination; the microlens array, configured to modulate a light beam with a preset angle to a target spatial position at a back focal plane of the microlens array to obtain a modulated image; an image sensor configured to record the modulated image; and a reconstruction module configured to reconstruct a 3D structure of the sample based on the modulated image.

Abstract: A computer-implemented method for processing images to identify Energy Infrastructure (EI) features within aerial images of global terrain is provided. The image processing method identifies information about EI features by applying an EI feature recognition model to aerial images of global terrain. The EI feature recognition model identifies the EI feature information according to image content of the aerial image. The method further provides updates to the identification of the EI feature information according to relationships between identified EI features.

Abstract: A cascading convolutional neural network (CCNN) comprising a plurality of convolutional neural networks (CNNs) that are trained by weighting training data based on loss values of each training datum between CNNs of the CCN. The CCNN can receiving an input image from plurality of images, classify the input image using the CCNN, and present a classification of the input image.

Abstract: Novel systems and methods for airway segmentation are disclosed using confident airway volume, multi-scale topological and geometric leakage detection, exact leakage localization and correction, forbidden volume insertion and freezing, iterative shift in growing techniques starting with a conservative parameter/strategy and progressing toward generous ones, and a final pruning using machine learning, neural network, deep learning or artificial intelligence methods. The systems and methods are fully automated requiring no manual inputs or post-editing steps. The systems and methods use region growing-based segmentation to iteratively grow an airway tree starting with an initial seed, possibly inside the trachea, and a conservative segmentation parameter, e.g., a conservative threshold value for intensity-based region growing. The method then gradually progresses toward more generous values of the segmentation parameters until there is convergence.

Abstract: The present disclosure relates to an image processing device and an image processing method for achieving an easy change of a display mode of each object in live-action content. The image processing device includes an image generation section that changes a display mode of each of objects within a display image on the basis of segment information that indicates a position of a segment in which each of the objects is present, the position of the segment being a position in each of a plurality of layer images that are images generated on the basis of a plurality of captured images and are images classified into a plurality of layers in accordance with distances of the images from a predetermined visual point. For example, the present disclosure is applicable to a display device or the like.

Abstract: A lung screening assessment system is operable to receive a chest computed tomography (CT) scan that includes a plurality of cross sectional images. Nodule classification data of the chest CT scan is generated by utilizing a computer vision model that is trained on a plurality of training chest CT scans to identify a nodule in the plurality of cross sectional images and determine an assessment score. A lung screening report that includes the assessment score of the nodule classification data is generated for display on a display device associated with a user of the lung screening assessment system.