Abstract: [Problem] To obtain high measurement accuracy in aerial photogrammetry while reducing the cost of installing a measurement target. [Solution] According to the present invention, a launching pad for an unmanned aircraft on which a UAV 200 is installed before take off is provided with: a target indication that constitutes a target locating point used for aerial photogrammetry; and position determining units that determine the position of the UAV 200 with respect to the launching pad in a state in which the position of the UAV 200 with respect to the target indication before take off is specified.

Abstract: A method and system for generating a combined rearview image of an area behind a vehicle that is towing a trailer, the method including: capturing a first image from a vehicle camera that is installed on the vehicle, the first image being comprised of an area behind the vehicle; capturing a second image from a trailer camera that is installed on the trailer, the second image being comprised of an area behind the trailer; determining an obstructed viewing region within the first image, the obstructed viewing region being a region of the first image in which the trailer resides; overlaying the second image on the first image and at least partially within the obstructed viewing region; and inserting graphics in the first image, in the second image, and/or in a patch region within the obstructed viewing region and between the second image and the first image.

Abstract: Systems and methods are provided for detecting an object in front of a vehicle. In one implementation, an object detecting system includes an image capture device configured to acquire a plurality of images of an area, a data interface, and a processing device programmed to compare a first image to a second image to determine displacement vectors between pixels, to search for a region of coherent expansion that is a set of pixels in at least one of the first image and the second image, for which there exists a common focus of expansion and a common scale magnitude such that the set of pixels satisfy a relationship between pixel positions, displacement vectors, the common focus of expansion, and the common scale magnitude, and to identify presence of a substantially upright object based on the set of pixels.

Abstract: A system for monitoring shopping carts uses cameras to generate images of the carts moving in a store. In some implementations, cameras may additionally or alternatively be mounted to the shopping carts and configured to image cart contents. The system may use the collected image data, and/or other types of sensor data (such as the store location at which an item was added to the basket), to classify items detected in the shopping carts. For example, a trained machine learning model may classify item in a cart as “non-merchandise,” “high theft risk merchandise,” “electronics merchandise,” etc. When a shopping cart approaches a store exit without any indication of an associated payment transaction, the system may use the associated item classification data, optionally in combination with other data such as cart path data, to determine whether to execute an anti-theft action, such as locking a cart wheel or activating a store alarm.

Abstract: A vehicular control system includes a plurality of cameras, a plurality of sensors and a central electronic control unit. Image data captured by at least one of the cameras and sensor data sensed by at least one of the sensors is provided to and processed at the central electronic control unit. Data relevant to a current geographical location of the vehicle is provided to and processed at the central electronic control unit. The central electronic control unit is operable to at least partially control the vehicle. The central electronic control unit may include a threat recognizer and a risk assessor. Responsive at least in part to threat recognition/evaluation by the threat recognizer and risk assessment by the risk assessor, the central electronic control unit may control at least one selected from the group consisting of (i) braking of the equipped vehicle and (ii) steering of the equipped vehicle.

Abstract: Systems and methods can be used for analyzing image data to determine an amount of vibration and/or misalignment in an object under analysis. In some instances, as operating equipment heats up during operation, temperature changes of various portions of the operating equipment leads to changes in dimensions of such portions, leading to misalignment. Multiple sets of data representative of the operating equipment in multiple operating conditions can be used to determine an amount of misalignment due to thermal offsets. Hot and cold temperatures of the equipment can be used to calculate thermal growth of various portions of the equipment, which can be used to determine an amount a misalignment due to thermal offsets. Additionally or alternatively, image data representing the equipment can be used to observe changes in alignment between states.

Abstract: An apparatus to facilitate real-time playback of point cloud sequence data is disclosed. The apparatus comprises one or more processors to receive point cloud data of a captured scene, decompose the point cloud data into a plurality of point cloud patches, wherein each point cloud patch is associated with an object in the scene and includes contextual information regarding the point cloud patch, encode each of the point cloud patches via a deep-learning based algorithm to generate encoded point cloud patches, receive a viewpoint selection from a client, assign a priority to data chunks within each encoded point cloud patch based on the viewpoint selection and the contextual information and transmit the data chunks to the client based on the assigned priority.

Abstract: The present invention relates to a method of performing motion compensation by using motion vector prediction. To this end, a method of decoding an image may include: generating multiple motion vector candidate lists according to an inter-prediction direction of a current block; deriving multiple motion vectors for the current block by using the multiple motion vector candidate lists; determining multiple prediction blocks for the current block by using the multiple motion vectors; and obtaining a final prediction block for the current block based on the multiple prediction blocks.

Abstract: The present invention relates to a method and device for sharing a candidate list. A method of generating a merging candidate list for a predictive block may include: producing, on the basis of a coding block including a predictive block on which a parallel merging process is performed, at least one of a spatial merging candidate and a temporal merging candidate of the predictive block; and generating a single merging candidate list for the coding block on the basis of the produced merging candidate. Thus, it is possible to increase processing speeds for coding and decoding by performing inter-picture prediction in parallel on a plurality of predictive blocks.

Abstract: A method for decoding a 360-degree image includes: receiving a bitstream obtained by encoding a 360-degree image; generating a prediction image by making reference to syntax information obtained from the received bitstream; combining the generated prediction image with a residual image obtained by dequantizing and inverse-transforming the bitstream to obtain a decoded image; and reconstructing the decoded image into a 360-degree image according to a projection format.

Abstract: A moving picture coding apparatus includes a co-located block information determination unit which determines which one of a forward reference block and a backward reference block will be a co-located block and further determines whether only the unidirectional motion vector of the motion vectors of the co-located block is to be stored in a colPic memory; a temporal motion vector predictor calculation unit which derives a candidate motion vector predictor in temporal motion vector predictor mode using the colPic information stored in the colPic memory; and an inter prediction control unit which determines to code a motion vector using a candidate motion vector predictor having least error from the motion vector derived by motion estimation among candidate motion vector predictors.

Abstract: Techniques for implementing a feature generation pipeline for machine learning are provided. In one technique, multiple jobs are executed, each of which computes a different set of feature values for a different feature of multiple features associated with videos. A feature registry is stored that lists each of the multiple features. After the jobs are executed and the feature registry is stored, a model specification is received that indicates a set of features for a model. For each feature in a subset of the set of features, a location is identified in storage where a value for said each feature is found and the value for that feature is retrieved from the location. A feature vector is created that comprises, for each feature in the set of features, the value that corresponds to that feature. The feature vector is used to train the model or as input to the model.

Abstract: A method for monitoring the surroundings of a vehicle, the surroundings behind the vehicle being detected with the aid of ultrasonic sensors and with the aid of at least one imaging sensor, a warning signal being output when an object that has a distance to the vehicle smaller than a predefined minimum distance is detected with the aid of ultrasonic sensors, a trailer of the vehicle being recognized with the aid of the imaging sensor, and the warning signal not being output when the object is the trailer recognized by the imaging sensor.

Abstract: An optical system for stereoscopic vision includes in order from an object side, a front unit and a rear unit. Each of the front unit and the rear unit includes a lens component consisting of a single lens or a cemented lens. The front unit includes a first front unit and a second front unit, and an optical axis of the first front unit, an optical axis of the second front unit, and an optical axis of the rear unit are positioned in a same plane. The optical axis of the rear unit is positioned between the optical axis of the first front unit and the optical axis of the second front unit, and the following conditional expression (1) is satisfied: 0.15<De/?<0.85??(1).

Abstract: A video transrater, and a transrater system and method. The transrater system and method may select an optimal bit rate from among a plurality of available bit rates in order to obtain a particular video quality and/or format.

Abstract: A three-dimensional scanning and tracking device to track at least one item, the three-dimensional scanning and tracking device including a main body, an imaging unit disposed on at least a portion of the main body, a display unit disposed on at least a portion of a center of the main body, a scan button disposed on at least a portion of the main body to scan a barcode of the at least one item using the imaging unit in response to depressing the scan button and orienting the imaging unit toward the barcode, and a camera button disposed on at least a portion of the main body to perform a three-dimensional image capture of the at least one item in response to depressing the camera button and orienting the imaging unit toward the at least one item, such that a three-dimensional image of the at least one item is created and displayed on the display unit.

Abstract: A method and system for generating a composite video for display in a vehicle. A plurality of video streams are generated from a plurality of video cameras configured to be positioned on the vehicle. The video streams are transformed by an electronic processor to create a virtual camera viewpoint. The transformed video streams are combined to generate a composite video including a portion of a first image that is generated from a first one of the video cameras. The electronic processor detects an object external to the vehicle and determines whether the object at least partially obscures the portion of the first image. When the object at least partially obscures the portion of the first image, the electronic processor supplements the portion of the first image with a portion of a second image that is generated by a second one of the video cameras.

Abstract: The invention describes an imaging device (1) comprising an image sensor (2), an imaging lens (3), an infrared light source (5) to illuminate a scene (SC), and an infrared autofocus system (6) for providing an autofocus function, wherein the image sensor (2) comprises an array (21) of sensor pixels each arranged as dual pixel (22) comprising two separate pixel sensors (22a, 22b) per dual pixel (22) to record the image data (D1) as a sum signal from both pixel sensors (22a, 22b) and providing infrared data (D2) as individual signals from each of the two pixel sensors (22a, 22b) for phase contrast (PC) autofocusing, wherein an infrared filter (7) arranged between an aperture (4) of the imaging device (1) and an imaging sensor (2) and is adapted to locally transmit only a portion of the infrared light (IR) to the imaging sensor (2) by comprising at least one first area (71) arranged as infrared blocking area and at least one second area (72) arranged as infrared bandpass area.

Abstract: A method includes capturing a first image of a scene, detecting a face in a section of the scene from the first image, and activating an infrared (IR) light source to selectively illuminate the section of the scene with IR light. The IR light source includes an array of IR light emitting diodes (LEDs). The method includes capturing a second image of the scene under selective IR lighting from the IR light source, detecting the face in the second image, and identifying a person based on the face in the second image.

Abstract: An exemplary embodiment may describe a convolutional explainable neural network. A CNN-XNN may receive input, such as 2D or multi-dimensional data, a patient history, or any other relevant information. The input data is segmented into various objects and a knowledge encoding layer may identify and extract various features from the segmented objects. The features may be weighted. An output layer may provide predictions and explanations based on the previous layers. The explanation may be determined using a reverse indexing mechanism (Backmap). The explanation may be processed using a Kernel Labeler method that allows the labelling of the progressive refinement of patterns, symbols and concepts from any data format that allows a pattern recognition kernel to be defined allowing integration of neurosymbolic processing within CNN-XNNs. The optional addition of meta-data and causal logic allows for the integration of connectionist models with symbolic logic processing.