Abstract: An information processing apparatus includes a setting unit configured to set an imaging condition under which an imaging apparatus captures a video, a region determination unit configured to determine a detectable region in which a detection target is detectable in the video, based on the imaging condition, an acquisition unit configured to acquire a desired detection condition under which a user desires detection for the detection target to be executed, and a condition determination unit configured to determine a detection condition under which the detection target is detected from the video, based on the desired detection condition and the detectable region determined based on at least one imaging condition.

Abstract: A system uses video of a vehicle to detect and classify the vehicle's pose. The system generates an image stack by scaling and shifting a set of digital image frames from the video to a fixed scale, yielding a sequence of images over a time period. The system processes the image stack with a classifier to determine the pose of the object. The system also may determine state and class of visible turn signals on the object, as well as predict the vehicle's direction of travel.

Abstract: A method for interactive image caricaturing by an electronic device is described. The method includes detecting at least one feature location of an image. The method further includes generating, based on the at least one feature location, an image mesh that comprises a grid of at least one horizontal line and at least one vertical line. The method additionally includes obtaining a gesture input. The method also includes determining at least one caricature action based on the at least one gesture input. The method further includes generating a caricature image based on the image mesh, the at least one caricature action and the image.

Abstract: Provided are embodiments for a method for performing colorization of scans. The method includes analyzing a scanner, a scan of an environment to identify one or more patterns within the scan, and obtaining a subset of colorization data of the environment. The method also includes predicting colors for the one or more patterns in the scan based on the subset of colorization data, and assigning the predicted colors to the one or more patterns in the scan to generate a colorized scan. The method includes displaying the colorized scan, wherein the colorized scan combines the scan and the predicted colorization data by assigning the predicted colorization data to the one or more patterns in the scan. Also provided are embodiments for a system for performing the colorization of scans.

Abstract: In some implementations, a service qualification system may receive, from a user device, a set of images that depict a window. The service qualification system may perform an image-based analysis of the set of images to determine a reflectivity score associated with the window. The reflectivity score may be indicative of a quality associated with a signal of the service being received through the window. The service qualification system may determine, based on the reflectivity score, a service qualification metric that is indicative of a capability of receiving the service within the unit. The service qualification system may perform an action associated with the service qualification metric.

Abstract: An information processing apparatus acquires a plurality of captured images captured by a plurality of image capturing devices, the plurality of captured images having imaging time codes that match each other, the plurality of captured images being used for generating a virtual viewpoint image, executes a detection process for detecting a predetermined image pattern on the plurality of captured images, and determines parameters on positions of the plurality of image capturing devices and orientations from the plurality of image capturing devices based on a result of the detection process for detecting the predetermined image pattern on the plurality of captured images having the imaging time codes that match each other.

Abstract: An apparatus which is capable of displaying an image for a user to easily recognize the brightness and colors of an area of interest in the image and easily determine whether a subject is in focus. The amount of image shift between parallax image signals is calculated. The amount of blur in an area where an image shift occurs in the parallax image signals is determined based on the amount of image shift. A blurring process is performed on at least one of the parallax image signals based on the amount of blur. An image based on the display image signal generated based on the resulting parallax image signal is displayed on a display. The determined amount of blur is greater than the amount of blur shown by a subject image defocused by the amount of defocus converted from the amount of image shift.

Abstract: Provided is an image processing device that can accurately detect a target object, even when a high-distortion lens is used. According to the present invention, a camera 100 captures images in accordance with a synchronization signal Sig1, a camera 101 captures images in accordance with a synchronization signal Sig2, an area of interest setting unit 1033 sets an area of interest that represents a region to which attention is to be paid, a phase difference setting unit 1034 sets a shift .DELTA.t (a phase difference) for synchronization signal Sig1 and synchronization signal Sig2 that synchronizes the imaging timing of camera 100 and camera 101 with respect to the area of interest in the images captured by camera 100 and a region of the images captured by camera 101 that corresponds to the area of interest, and a synchronization signal generation unit 102 generates synchronization signal Sig1 and synchronization signal Sig2 on the basis of the shift .DELTA.t.

Abstract: A series of smart phones are mounted in respective tripods to capture motion of a person wearing markers, such as marker balls or reflectors. The videos from the phones are stripped of objects other than the markers and the videos of the markers are combined to render a 3D motion capture structure that may be applied to an image of a VR icon to cause the VR icon to move as the person originally moved.

Abstract: Provided is a control apparatus that receives information regarding a motion of user's fingers and information regarding a shape of user's hands, generates first pose information indicative of a pose of user's hands in accordance with the received information regarding the motion of user's fingers, generates second pose information indicative of a pose of user's hands in accordance with the received information regarding the shape of user's hands, combines the generated first pose information with the generated second pose information in accordance with criteria based on a predetermined condition, and generates and outputs information indicative of the pose of user's hands.

Abstract: The present disclosure relates to an image processing apparatus, an image processing method, and a program that permit acquisition of a more natural-looking synthesized image. A first captured image depicting a first subject is separated into a foreground area and a background area. The first subject appears in the foreground area. The background area is an area other than the area where the first subject appears. Also, a blended image is generated to be displayed on a live view, in which the blended image is obtained by blending the foreground area and the background area of the first captured image with an output image that is output when a second subject is captured. Then, a synthesized image is generated by superimposing the foreground area of the first captured image on the second captured image depicting the second subject with reference to the blended image in such a manner that the backgrounds are substantially identical. The present technology is applicable, for example, to an imaging apparatus.

Abstract: Disclosed are techniques for estimating a 3D bounding box (3DBB) from a 2D bounding box (2DBB). Conventional techniques to estimate 3DBB from 2DBB rely upon classifying target vehicles within the 2DBB. When the target vehicle is misclassified, the projected bounding box from the estimated 3DBB is inaccurate. To address such issues, it is proposed to estimate the 3DBB without relying upon classifying the target vehicle.

Abstract: A visual and geolocation analytic system is provided, including: an analytic device and a number of image capturing devices connected to said analytic device. The image capturing devices capture images of an object at a time interval and send said captured images to said analytic device; said analytic device comprises a deep learning model for analyzing said captured images, allowing said object to be identified and tagged, and allowing a path of movement of said object across time to be tracked. The present invention tracks the position of an object within an area continuously across time, and transform the object in captured images into structured data set for analysis.

Abstract: Systems and methods for more accurate and robust determination of subject characteristics from an image of the subject. One or more machine learning models receive as input an image of a subject, and output both facial landmarks and associated confidence values. Confidence values represent the degrees to which portions of the subject's face corresponding to those landmarks are occluded, i.e., the amount of uncertainty in the position of each landmark location. These landmark points and their associated confidence values, and/or associated information, may then be input to another set of one or more machine learning models which may output any facial analysis quantity or quantities, such as the subject's gaze direction, head pose, drowsiness state, cognitive load, or distraction state.

Abstract: Methods, devices, systems and apparatus for determining emphysema thresholds for processing a pulmonary medical image are provided. In one aspect, a method includes: determining lung lobe regions in the pulmonary medical image, and, for each of the lung lobe regions, clustering CT values in the lung lobe region to divide the lung lobe region into a first sub region and a second sub region and acquiring a CT value corresponding to an intersection of a first CT value distribution function for the first sub region and a second CT value distribution function for the second sub region in the lung lobe region as an emphysema threshold for the lung lobe region.

Abstract: An image processing part applies an image process by software to a surgical region image, and a display control part controls a display of the surgical region image to which the image process is applied. The image processing part produces a pre-update processed image acquired by applying the image process established before updating the software to the surgical region image and a post-update processed image acquired by applying the image process established after updating the software to the surgical region image, and the display control part controls a display of at least a portion of at least either one of the pre-update processed image or the post-update processed image. The present technique is applicable to a CCU of an endoscopic surgery system.

Abstract: An information processing apparatus includes a scene obtaining unit and a suitability determining unit. The scene obtaining unit is configured to obtain a scene including position and attitude information of a movable object equipped with a camera and information regarding a three-dimensional shape of an object to be imaged, the object to be imaged being imaged by the camera. The suitability determining unit is configured to determine suitability of the scene on the basis of suitability data indicating whether or not the obtained scene is suitable for localizing processing.

Abstract: The present technology relates to an information processing device, an information processing method, a program, and an information processing system for enabling reduction in a load by enabling selective arithmetic processing in performing imaging without using an imaging lens. The information processing device includes an acquisition unit configured to acquire a detection image output from an imaging element that receives incident light incident without through an imaging lens, and restoration information including setting information set by a user and to be used to generate a restoration image from the detection image, a restoration processing unit configured to perform restoration processing of generating the restoration image using the detection image and the restoration information, and an output control unit configured to control an output of the restoration image. The present technology can be applied to, for example, a device or a system that restores a detection image captured by a lensless camera.

Abstract: A method of image-based analysis of multiple samples includes using a sample holder having multiple locations of interest and multiple focal structures that are each associated, one or more, with the multiple locations of interest, wherein the multiple samples are dispersed across the multiple locations of interest and obtaining image areas of the multiple locations of interest. Multiple digital image areas are thus obtained for use in an analysis of the multiple samples with each of the image areas including at least one of the locations of interest and at least one of the focal structures. An image processing algorithm is used to analyse each of the digital image areas and check if the focal structure indicates that the image area is in clear focus. An indication is provided and/or remedial action is taken if the image processing algorithm indicates that any digital image areas are out of focus.

Abstract: A method and device for image processing, a method for training an object detection model are provided in the present disclosure. In the method for image processing, a visible light image is acquired. A central weight map corresponding to the visible light image is generated. Weight values represented by the central weight map gradually decrease from a center to an edge of a visible light image. The visible light image and the central weight map are inputted into an object detection model to obtain an object region confidence map. The object detection model is a model obtained by training according to multiple set of training data, each set of which includes a visible light image, a central weight map and a corresponding labeled object mask pattern for a same scenario. A target object in the visible light image is determined according to the object region confidence map.