Abstract: In one embodiment, an apparatus includes accessing an image, captured by a first camera, of an object that includes a size-invariant feature and a size-variant feature and determining a size of the size-invariant feature in the image. The method further includes determining, based on the determined size of the size-invariant feature in the image, a distance between the camera and the object and determining a size of the size-variant feature in the image. The method further includes determining, based on the determined distance between the camera and the object and based on the determined size of the size-variant feature in the image, an actual size of the size-variant feature.

Abstract: A gesture recognition device that recognizes, based on an image obtained by imaging a person to be imaged with a digital camera having an imaging range changing mechanism for changing an imaging range, a gesture represented by a hand of the person to be imaged includes at least one processor. The at least one processor controls, in a case where a position of a face of the person to be imaged in the image deviates from a predetermined set position, an operation of the imaging range changing mechanism to cause the digital camera to capture a proper face position image with the position of the face as the set position, and determines whether the hand detected from the proper face position image is a right hand or a left hand based on a positional relationship between a position of the detected hand and the set position.

Abstract: Some embodiments of the present disclosure provide methods, devices, computer-readable storage mediums for detecting a signal light. The method may include obtaining a first image and a second image previous to the first image in time sequence, wherein both the first image and the second image include a same target signal light; determining, based on the first image, a first state of the target signal light in the first image; determining, based on the second image and the first image, a second state of the target signal light in the first image; and determining, based on the first state and the second state, a target state of the target signal light at a time point when the first image is captured.

Abstract: An apparatus obtains a mesh representation of a 3D object, and performs an encoding process: segmenting the 3D object into 2D projections; encoding the segmented 2D projections into a bitstream; reconstructing a 3D mesh from the segmented and encoded 2D projections; deriving an error vector, describing a shift between 3D data, from patches reconstructed from corresponding encoded patches of the bitstream, and original data in the mesh representation. The patches are submeshes of the reconstructed 3D mesh. The apparatus signals and transmits the error vector in or along with the bitstream. A decoder apparatus receives these and performs a decoding process to form decoded 2D projections of the mesh representation, reconstructs a 3D mesh from decoded 2D projections, applies the patch correction vector elements from the error vector in the decoding process or the reconstructing, produces a modified 3D mesh, as an output mesh and outputs the output mesh.

Abstract: In an aspect, an image processing apparatus includes a processor and a memory in which a plurality of images obtained by capturing images of a building and a three-dimensional model of the building are stored in association with each other. The processor is configured to perform a development process to perform development of the three-dimensional model into a two-dimensional image, an extraction process to extract defect information of the building on the basis of the plurality of images, a mapping process to perform mapping of the defect information to the two-dimensional image, and an output process to output the two-dimensional image to which the mapping is performed.

Abstract: A method includes receiving a first set of coordinates associated with a first geostationary point, identifying a first plurality of geostationary points having a clear line of sight to the first geostationary point and a second plurality of points lacking the clear line of sight to the first geostationary point, and generating a map in which the first plurality of geostationary points is depicted in a first visual style. Optionally, the map may further depict the second plurality of geostationary points in a second visual style different from the first visual style.

Abstract: In some implementations, a method includes: obtaining image data associated with a physical environment; obtaining first contextual information including at least one of first user information associated with a current state of a user of the computing system, first application information associated with a first application being executed by the computing system, and first environment information associated with a current state of the physical environment; selecting a first set of perspective correction operations based at least in part on the first contextual information; generating first corrected image data by performing the first set of perspective correction operations on the image data; and causing presentation of the first corrected image data.

Abstract: A control method and a control system for image scanning, an electronic apparatus, and a storage medium. The control method for image scanning comprises of: configuring a monitoring-layer scanning, and obtaining monitoring-layer images; determining an initial target region in each of the monitoring-layer images, and a target part being located in the initial target region; identifying the target part in the initial target region, and determining a region corresponding to the identified target part to be a region of interest (ROI); and generating a computed tomography (CT) control parameter based on a CT number of the ROI, and adjusting a trigger time of image scanning according to the CT control parameter.

Abstract: A method for lane detection, the method includes (a) determining, by a lane detection neural network (NN) and based on locations of one or more road lanes within an aerial image, locations of the one or more road lanes on a ground vehicle acquired (GVA) image; and (b) responding to the determining of the locations of the one or more road lanes on the GVA image. The lane detection NN was trained to perform aerial-images-to-GVA-images conversion. The aerial image was generated by a first NN, based in the GVA image. The first NN was trained to perform GVA-images-to-aerial-images conversion. The GVA image was acquired by a two dimensional camera of a ground vehicle. The one or more locations of the one or more road lanes within the aerial image are provided by a second NN that was trained to perform lane detection in arial images.

Abstract: A super-resolution reconstruction method for adaptive streaming of media over a network includes streaming an instance of streamable media to a terminal device via the network, identifying a streaming resolution of the instance of the streamable media based on the streaming resolution being less than a first target super-resolution, acquiring, by the terminal device from a server of the network, a first target super-resolution neural network model corresponding to the identified streaming resolution and the first target super-resolution, reconstructing the instance of the streamable media streamed to the terminal device into an instance of the streamable media at the first target super-resolution using the first target super-resolution neural network model and playing the reconstructed instance of the streamable media by the terminal device, wherein the first target super-resolution neural network model is obtained by performing model training in advance by taking a media at the identified streaming resolutio

Abstract: Three-dimensional data acquisition means acquires three-dimensional data output from a three-dimensional ranging sensor for scanning a paved surface for aircraft using laser light. Extraction means extracts, from the three-dimensional data, real image data about a real image present on the paved surface and mirage image data about a mirage image corresponding to the real image. Foreign object detection means detects a foreign object left on the paved surface using the real image data and the mirage image data.

Abstract: A system and method are provided for automatically extracting image data of a driver's face when a driver sits in a driver's seat and then starting an engine of a vehicle, based on determining whether the extracted face matches a pre-registered face, and automatically controlling a vehicle environment based on the matched face data, thereby maximizing the convenience of using the vehicle.

Abstract: Approaches presented herein provide for the maintaining of fine details that might be removed by a denoiser used to reduce an amount of noise in an image. An input image can be provided to a denoiser, and can also can be simultaneously processed to extract pixel data that may correspond to fine detail or high frequency features. Individual pixels of an image can have a value determined for a material property sampled for that pixel location, and that value can be compared against an average material property value determined for neighboring pixels. The ratio of material values can be multiplied by the value of a corresponding pixel of the denoised image, for any or all pixel locations, to obtain final pixel values for an output image that include less noise than the original image but represent fine detail that may otherwise have been lost during the denoising process.

Abstract: In various examples, determining point bounding shapes for systems and applications is described herein. Systems and methods are disclosed that determine a bounding shape (e.g., a minimum-area bounding rectangle) for an object and/or a point set associated with the object. In some examples, to determine the bounding shape, the systems and methods may rotate points from the point set to include various orientations. The systems and methods may then determine a respective bounding shape for one or more (e.g., each) orientation of the points and use the bounding shapes to determine a final bounding shape for the points. For example, the bounding shape that is associated with the smallest area may be selected for the points. The bounding shape may then be rotated by an angle in order to determine a final bounding shape for the object.

Abstract: Provided are systems and methods to perform colour extraction from swatch images and to define new images using extracted colours. Source images may be classified using a deep learning net (e.g. a CNN) to indicate colour representation strength and drive colour extraction. A clustering classifier is trained to use feature vectors extracted by the net. Separately, pixel clustering is useful when extracting the colour. Cluster count can vary according to classification. In another manner, heuristics (with or without classification) are useful when extracting. Resultant clusters are evaluated against a set of (ordered) expected colours to determine a match. Instances of standardized swatch images may be defined from a template swatch image and respective extracted colours using image processing. The extracted colour may be presented in an augmented reality GUI such as a virtual try-on application and applied to a user image such as a selfie using image processing.

Abstract: A method comprising: obtaining a plurality of data sets, each of the plurality of data sets being generated, at least in part, by using by a different one of a plurality of sensors, each of a plurality of sensors including a wearable sensor that is worn by a user or a sensor that is positioned at a same location as the user; combining the plurality of data sets to produce a fused data set; processing the fused data set to identify at least one of a performance of the user in completing a task, a cognitive load of the user, and/or one or more objects that are positioned at the same location as the user; and outputting an indication of a state of the user based on an outcome of the processing.

Abstract: A method of conditional neural ground planes for static-dynamic disentanglement is described. The method includes extracting, using a convolutional neural network (CNN), CNN image features from an image to form a feature tensor. The method also includes resampling unprojected 2D features of the feature tensor to form feature pillars. The method further includes aggregating the feature pillars to form an entangled neural ground plane. The method also includes decomposing the entangled neural ground plane into a static neural ground plane and a dynamic neural ground plane.

Abstract: A method of providing information for predicting the prognosis of breast reconstruction surgery and a system for predicting the prognosis of breast reconstruction surgery is provided. By using the method or system for providing information for predicting the prognosis of breast reconstruction surgery according to one specific example of the present invention, it is possible to predict the possibility of implant rupture in a subject who has undergone or is scheduled to undergo two-stage breast reconstruction surgery, and it is possible to prepare an appropriate treatment according to this prediction, and to adjust the interval between breast reconstruction surgeries so as to lower the possibility of implant rupture. In addition, it is possible to provide information so that an implant with a low possibility of rupture may be selected.

Abstract: Systems and methods for altering colors of pixels of images produced by cameras using a lens. A plurality of generated images are received, and color channel values and luminance values of image pixels are determined. Benchmark U-color thresholds and V-color thresholds are established based on the color channels and luminance associated with the pixels. A first image generated using a first lens of a first camera is received. U-color values and V-color values associated with first pixels of the first image are determined. A uniformity score associated with the first lens is received. The benchmark U-color and V-color thresholds are altered based on the uniformity score. The U-color values and V-color values of the first pixels are corrected in response to the U-color values being within the altered U-color thresholds and the V-color values being within the altered V-color thresholds.

Abstract: Implementations are disclosed for analyzing extracted individual plant components to make agronomic inferences about entire composite plant organs from which the individual plant components were harvested, and for using those agronomic inferences for various purposes. In various implementations, individual plant component(s) may be sampled from multiple plant components removed from composite plant organ(s) that previously included the plant components. Digital image(s) may be captured of the sampled individual plant component(s) and processed based on machine learning model(s) to generate agronomic inference(s) about the composite plant organ(s) that previously included the plurality of plant components. Based on the agronomic inference(s), computing device(s) may render output that includes a diagnosis or recommendation for the grower about the field or the crops, and/or agricultural equipment (e.g., robots) may be operated automatically.