Erroneous detection due to erroneous parallax measurement is suppressed to accurately detect a step present on a road. An in-vehicle environment recognition device 1 includes a processing device that processes a pair of images acquired by a stereo camera unit 100 mounted on a vehicle. The processing device includes a stereo matching unit 200 that measures a parallax of the pair of images and generates a parallax image, a step candidate extraction unit 300 that extracts a step candidate of a road on which the vehicle travels from the parallax image generated by the stereo matching unit 200, a line segment candidate extraction unit 400 that extracts a line segment candidate from the images acquired by the stereo camera unit 100, an analysis unit 500 that performs collation between the step candidate extracted by the step candidate extraction unit 300 and the line segment candidate extracted by the line segment candidate extraction unit 400 and analyzes validity of the step candidate based on the collation result and an inclination of the line segment candidate, and a three-dimensional object detection unit 600 that detects a step present on the road based on the analysis result of the analysis unit 500.

This application provides an image processing method, a network training method, and a related device, and relates to image processing technologies in the artificial intelligence field. The method includes: inputting a first image including a first vehicle into an image processing network to obtain a first result output by the image processing network, where the first result includes location information of a two-dimensional 2D bounding frame of the first vehicle, coordinates of a wheel of the first vehicle, and a first angle of the first vehicle, and the first angle of the first vehicle indicates an included angle between a side line of the first vehicle and a first axis of the first image; and generating location information of a three-dimensional 3D outer bounding box of the first vehicle based on the first result.

A method for calculating a distance between a vehicle camera and an object, the method may include: (a) obtaining an image that was acquired by the vehicle camera of a vehicle; the image captures the horizon, the object, and road lane boundaries; (b) determining an initial row-location horizon estimate and a row-location contact point estimate, the contact point is between the object and a road on which the vehicle is positioned; (c) determining a vehicle camera roll angle correction that once applied will cause the lanes boundaries to be parallel to each other in the real world; (d) calculating a new row-location horizon estimate, wherein the calculating comprises updating the row-location horizon estimate based on the vehicle camera roll angle correction; and (e) calculating the distance between the vehicle camera based on a difference between the new row-location horizon estimate and the row-location contact point estimate.

The described aspects and implementations enable fast and accurate verification of radar detection of objects in autonomous vehicle (AV) applications using combined processing of radar data and camera images. In one implementation, disclosed is a method and a system to perform the method that includes obtaining a radar data characterizing intensity of radar reflections from an environment of the AV, identifying, based on the radar data, a candidate object, obtaining a camera image depicting a region where the candidate object is located, and processing the radar data and the camera image using one or more machine-learning models to obtain a classification measure representing a likelihood that the candidate object is a real object.

In various examples, a sequential deep neural network (DNN) may be trained using ground truth data generated by correlating (e.g., by cross-sensor fusion) sensor data with image data representative of a sequences of images. In deployment, the sequential DNN may leverage the sensor correlation to compute various predictions using image data alone. The predictions may include velocities, in world space, of objects in fields of view of an ego-vehicle, current and future locations of the objects in image space, and/or a time-to-collision (TTC) between the objects and the ego-vehicle. These predictions may be used as part of a perception system for understanding and reacting to a current physical environment of the ego-vehicle.

A system for method for future forecasting using action priors that include receiving image data associated with a surrounding environment of an ego vehicle and dynamic data associated with dynamic operation of the ego vehicle. The system and method also include analyzing the image data and detecting actions associated with agents located within the surrounding environment of the ego vehicle and analyzing the dynamic data and processing an ego motion history of the ego vehicle. The system and method further include predicting future trajectories of the agents located within the surrounding environment of the ego vehicle and a future ego motion of the ego vehicle within the surrounding environment of the ego vehicle.

Systems and methods for navigating a host vehicle. The system may perform operations including receiving, from an image capture device, at least one image representative of an environment of the host vehicle; analyzing the at least one image to identify an object in the environment of the host vehicle; determining a location of the host vehicle; receiving map information associated with the determined location of the host vehicle, wherein the map information includes elevation information associated with the environment of the host vehicle; determining a distance from the host vehicle to the object based on at least the elevation information; and determining a navigational action for the host vehicle based on the determined distance.

A vehicular trailer hitching assist system includes a rear backup camera disposed at a rear portion of a vehicle and a control disposed in the vehicle. A display device includes a video display screen operable to display video images for viewing by a driver of the vehicle. During a hitching maneuver event undertaken by the driver to hitch a trailer hitch of the vehicle to a trailer tongue of a trailer, and while the driver of the vehicle is maneuvering the vehicle to hitch the vehicle to the trailer, the control outputs video images derived from image data captured by the rear backup camera for display by the video display screen. At least one overlay overlays the displayed video images to guide the driver during the hitching maneuver. The at least one overlay aids in guiding connection of the trailer hitch of the vehicle to the trailer tongue of the trailer.

The offline map generation process may collect multiple point cloud data of the same area. A perception algorithm may operate on the point cloud data to detect static objects, which may be fixed road features that do not change among the point cloud data, allowing the perception algorithm to more accurately detect the static objects. During online operation of the ADV through the area, the ADV may trim regions-of-interest (ROI) of the area to exclude the predefined static objects. The perception algorithm may execute the sensor data of the ROI in real-time to detect objects in the ROI. The may be added back to the output of the perception algorithm to complete the perception output.

Disclosed are an apparatus and a method for controlling a lane change using V2V communication information and an apparatus for calculating tendency information for the same. According to the apparatuses and the method, it is possible to improve safety when changing lanes by receiving diving information of drivers of other vehicles from communication modules of the other vehicles, generating tendency information of the drivers of the other vehicles on the basis of the driving information, and performing lane change control using the tendency information.