Abstract: A method for classifying objects on a roadway in surroundings of a vehicle. The method includes: reading in image data from a vehicle camera of the vehicle. The image data represent an area of the surroundings which includes the roadway; evaluating the image data including generating a model of a surface of the roadway using identified roadway markings, and an object on the roadway being identified; ascertaining first distance values between the vehicle camera and object image points of the object represented by the image data, and second distance values between the vehicle camera and roadway image points, defined by the model, of the surface of the roadway in surroundings of the object; and comparing the distance values to at least one continuity criterion for distinguishing raised objects from flat objects to classify the object as a raised or flat object as a function of a result of the comparison.

Abstract: An information provision device includes a controller configured to detect installation of an anti-skid tool onto a vehicle from an image including at least a part of the vehicle and at least a part of the vicinity of the vehicle and a communication unit configured to provide position information of the vehicle at the time of the installation of the anti-skid tool detected by the controller. A driving assistance device includes a communication unit configured to acquire position information of a vehicle at the time of installation of an anti-skid tool when the installation of the anti-skid tool onto the vehicle is detected from an image including at least a part of the vehicle and at least a part of a vicinity of the vehicle, and a controller configured to present information acquired by the communication unit to the driver of another vehicle different from the vehicle.

Abstract: An image-based road cone recognition method, apparatus, storage medium, and vehicle. Said method comprises: acquiring, during vehicle driving, an image of an object to be recognized; performing differential processing of the image, so as to acquire an image on which the differential processing has been performed, and performing, according to a preset threshold, ternary processing of the image on which the differential processing has been performed, so as to acquire a ternary image comprising forward boundary pixels and negative boundary pixels; acquiring, according to the forward boundary pixels and the negative boundary pixels, a forward straight line segment and a negative straight line segment which represent the trend of the boundaries of the object to be recognized; when position information of the forward and negative straight line segments matches boundary position information of a known road cone, determining that the object to be recognized is a road cone.

Abstract: The invention enables spoof detection in eye based person detection, person recognition, or person monitoring systems. The invention includes (i) illuminating an eye from a first source located at a first position, (ii) illuminating the eye from a second source located at a second position spaced from the first position, (iii) acquiring at the image sensor, a set of images of the eye which includes (a) a first set of image information representing a first specular reflection at a third position relative to the eye, (b) a second set of image information representing a second specular reflection at a fourth position relative to the eye, (iv) determining a difference value representing a difference between the third position and the fourth position, and (v) generating a data signal representing detection of a real eye in response to determining that the difference value is less than a threshold value.

Abstract: An apparatus and a method for detecting, classifying and tracking road users on frames of video data are provided. An object detection unit identifies an object in a frame of the video data. A pre-filtering unit determines whether the identifier associated to an object corresponds to another identifier associated to the same object in the same frame of the video data when the identifiers overlap. A tracking unit matches the identified object to a corresponding object in another frame of the video data, such that a path of the sequence of matched objects is tracked throughout the frames of the video data. A scenario unit determines a safety indicator that is a function of the risk indicators indicating interactions between road users.

Abstract: To make error correction in a position estimate of a vehicle, visual lane markings on the road can be matched with lane boundaries for the road within 3-D map data. Embodiments include obtaining a “stripe” indicative of an observed lane marking captured in a from a camera image, obtaining map data for the road on which the vehicle is located, determining the plane of the road from coordinates of lane boundaries within the map data, projecting the stripe onto the plane, and comparing the projected stripe with lane boundaries within the map data and associating the visual lane markings with the closest lane boundary. Differences between the projected stripe and the associated lane boundary can then be used for error correction in the position estimate of the vehicle.

Abstract: A method of predicting a trajectory of a target vehicle with respect to an ego vehicle includes receiving a first image having a target object representing a target vehicle on a road from a camera mounted on the ego vehicle, generating a bounding box associated with the target vehicle, the bounding box surrounding a contour of the target object, extracting tire contact information of the target object with a road surface from the first image, wherein each tire contact information represents coordinates, in the first image, of contact between a corresponding tire and the road, mapping the plurality of tire contact information from the first image to a bird's eye view image of the first image, and calculating the trajectory, with reference to a moving direction of the ego vehicle, of the target object using the mapped tire contact information on the bird's eye view image.

Abstract: Disclosed is a statistical method, apparatus and device for customer flow. The method includes: acquiring video data on which the statistics is to be made (S101), identifying facial areas in the video data (S102), matching the identified facial areas with pre-set facial information, wherein the pre-set facial information may be facial information of a person unrelated to the customer flow (for example, a staff, etc.); determining the number of successfully matched facial areas, to obtain the customer flow without unrelated persons (S103). It can be seen that the method removes the interference from unrelated persons and improves the accuracy of customer flow statistics.

Abstract: The present disclosure relates to systems and methods for road edge detection and mapping, for vehicle wheel identification and navigation based thereon, and for classification of objects as moving or non-moving. Such systems and methods may include the use of trained systems, such as one or more neural networks. Further, autonomous vehicle systems may incorporate aspects of one or more of the disclosed systems and methods.

Abstract: An image processor according to the present disclosure includes: a multiplier that receives image data from a pixel section including pixels of a plurality of colors and multiplies the image data by an adjustment parameter that adjusts a color level in each of the pixels; an adjuster that calculates a ratio of respective colors in each of the pixels in the image data and adjusts a value of the adjustment parameter on the basis of the ratio of the respective colors; and a binarization processor that extracts a target image of a specific color on the basis of the image data multiplied by the adjustment parameter.

Abstract: An integrated facial recognition method and system. The method includes: receiving a request for acquiring an integrated facial recognition service sent by a user terminal, which includes: an identifier of a user-selected model associated with facial recognition of the user, and an identifier of an operation selected by the user from candidate operations; and executing distributedly an operation selected by the user from the candidate operations on the user-selected model associated with the facial recognition of the user to obtain an operation result, and storing the operation result. The embodiment has realized completing the operations such as training a model or developing a facial recognition application, without the need of buying hardware and establishing a software environment by the user, thereby saving the development cost and improving the convenience of using the facial recognition service.

Abstract: A computer, including a processor and a memory, the memory including instructions to be executed by the processor to determine a plurality of topological nodes wherein each topological node includes a location in real-world coordinates and a three-dimensional point cloud image of the environment at the location of the topological node and process an image acquired by a sensor included in a vehicle using a variational auto-encoder neural network trained to output a semantic point cloud image, wherein the semantic point cloud image includes regions labeled by region type and region distance relative to the vehicle.

Abstract: A dynamic image analysis apparatus includes a hardware processor that acquires an X-ray dynamic image including continuous frame images acquired by continuously capturing a living body having a heartbeat in time series; performs logarithmic conversion for a pixel value of the acquired X-ray dynamic image to create a logarithmically converted image; sets, as a reference frame image, one frame image based on a heartbeat phase in at least one of the X-ray dynamic image and the logarithmically converted image; calculates (i) a difference or ratio between the X-ray dynamic image as the reference frame image and the X-ray dynamic image as a comparative frame image which is another frame image or (ii) a difference or ratio between the logarithmically converted image as the reference frame image and the logarithmically converted image as the comparative frame image; and generates a blood flow analysis image.

Abstract: An object detection apparatus 100 is an apparatus for detecting an object in a fish-eye image. The object detection apparatus 100 includes a normalized image acquisition unit 10 configured to acquire a normalized image obtained by normalizing a fish-eye image in which an object appears; a position detecting unit 20 configured to detect position coordinates of the object in the normalized image; and a determination unit 30 configured to determine a positional relationship between the object and the object detection apparatus 100 using the position coordinates of the object in the normalized image.

Abstract: A misregistration metrology system and method useful in the manufacture of semiconductor devices, the multilayered semiconductor devices including a first periodic structure having a first pitch along a first axis, the first periodic structure being formed together with a first layer of the multilayered semiconductor device, a second periodic structure having a second pitch along a second axis, the second axis not being parallel to the first axis, the second periodic structure being formed together with the first layer of the multilayered semiconductor device and a third periodic structure having a third pitch along a third axis, the third axis not being parallel to the first axis and the third axis not being parallel to the second axis, the third periodic structure being formed together with a second layer of the multilayered semiconductor device, the third periodic structure and the first and second periodic structures overlying one another, the misregistration metrology system and method including generati

Abstract: Various embodiments describe video object segmentation using a neural network and the training of the neural network. The neural network both detects a target object in the current frame based on a reference frame and a reference mask that define the target object and propagates the segmentation mask of the target object for a previous frame to the current frame to generate a segmentation mask for the current frame. In some embodiments, the neural network is pre-trained using synthetically generated static training images and is then fine-tuned using training videos.

Abstract: Provided are a method for setting an owner of an object based on sensing information regarding an internal environment of a vehicle when the object comes apart from a passenger, and an electronic apparatus therefor. In the present disclosure, one or more of an electronic apparatus, a vehicle, a vehicular terminal, and the autonomous driving vehicle may be associated with an artificial intelligence module, an unmanned aerial vehicle (UAV), a robot, an augmented reality (AR) device, a virtual reality (VR) device, a 5G service-related device, and the like.

Abstract: Described herein are methods and systems for keyframe-based object scanning and tracking. A sensor device captures images of objects in a scene. For each image, a computing device labels each of at least a plurality of pixels in the image, tracks at least one region of the labeled image to determine an estimate of a current pose of at least one object, validates the estimate of the current pose of the at least one object, selects the labeled image as a keyframe based upon validation of the estimate of the current pose, and updates a volumetric model comprising the at least one object using the keyframe. The computing device generates a final 3D model of the at least one object based upon the updated volumetric model.

Abstract: Methods and systems for generating and utilizing a road friction estimate (RFE) indicating the expected friction level between a road surface and the tires of a vehicle based on forward looking camera image signal processing. A forward-looking camera image is pre-processed, patch segmented (both laterally and longitudinally, as defined by wheel tracks or the like), transformed into a bird's eye view (BEV) image using perspective transformation, patch quantized, and finally classified. The resulting RFE may be used to provide driver information, automatically control the associated vehicle's motion, and/or inform a cloud-based alert service to enhance driver safety.

Abstract: Embodiments of the present disclosure relate to object defect detection. In an embodiment, a computer-implemented method is disclosed. According to the method, for a test image of at least one part of a target object, a reference image is generated by repeating a periodic pattern detected in the test image, the target object consisting of elements. A differential image is determined by comparing the test image and the reference image. The differential image is superimposed on a predefined grid image to obtain a superimposed image. The grid image comprises grids corresponding to elements of a reference object associated with the target object. The number of defective elements is determined in the at least one part of the target object based on the superimposed image. In other embodiments, a system and a computer program product are disclosed.