The invention belongs to motion detection and three-dimensional image analysis technology fields. It can be applied to movement detection or intrusion detection in the surveillance of monitored volumes or monitored spaces. It can also be applied to obstacle detection or obstacle avoidance for self-driving, semi-autonomous vehicles, safety systems and ADAS. A three-dimensional imaging system stores 3D surface points and free space locations calculated from line of sight data. The 3D surface points typically represent reflective surfaces detected by a sensor such as a LiDAR, a radar, a depth sensor or stereoscopic cameras. By using free space information, the system can unambiguously derive a movement or an intrusion the first time a surface is detected at a particular coordinate. Motion detection can be performed using a single frame or a single 3D point that was previously a free space location.

Systems and methods are provided for controlling vehicle operation. A processor may access route information for navigation of a route by the vehicle including data relating to speed along the route and calculate a speed of the vehicle along the route based on the route information. The processor may cause the vehicle to be operated at the calculated speed along the route; obtain dynamic information for the route based on data collected from one or more other vehicles on the route and indicating current conditions on the route which affect the speed of the vehicle along the route; and cause the vehicle to be operated at an updated speed along the route, based on the dynamic information.

The present disclosure relates to a method for generating a view for a camera system, in particular a surround-view camera system for a vehicle, including a control device and at least one camera, wherein the view is generated by means of the following method steps: capturing at least one object from the environment data from the at least one camera; generating a bounding box for the object; projecting the object onto a ground plane; creating a bounding shape which includes the bounding box and the projected object; creating a mesh structure or grid structure for the bounding shape; and arranging the mesh structure or grid structure within the bounding box, wherein the bounding shape is adapted, in particular by image scaling and/or image distortion, to the size of the bounding box.

The technology relates to controlling a vehicle in an autonomous driving mode. For instance, sensor data identifying an object in an environment of the vehicle may be received. A grid including a plurality of cells may be projected around the object. For each given one of the plurality of cells, a likelihood that the object will enter the given one within a period of time into the future is predicted. A contour is generated based on the predicted likelihoods. The vehicle is then controlled in the autonomous driving mode in order to avoid an area within the contour.

A system and method for estimating free space including applying a machine learning model to camera images of a navigation area, where the navigation area is broken into cells, synchronizing point cloud data from the navigation area with the processed camera images, and associating probabilities that the cell is occupied and object classifications of objects that could occupy the cells with cells in the navigation area based on sensor data, sensor noise, and the machine learning model.

A system is comprised of one or more processors coupled to memory. The one or more processors are configured to receive image data based on an image captured using a camera of a vehicle and to utilize the image data as a basis of an input to a trained machine learning model to at least in part identify a distance of an object from the vehicle. The trained machine learning model has been trained using a training image and a correlated output of an emitting distance sensor.

A system (1) for detecting dynamic secondary objects (55) that have a potential to intersect the trajectory (51) of a moving primary object (50), comprising a vision sensor (2) with a light-sensitive area (20) that comprises event-based pixels (21), so that a relative change in the light intensity impinging onto an event-based pixel (21) of the vision sensor (2) by at least a predetermined percentage causes the vision sensor (2) to emit an event (21a) associated with this event-based pixel (21), wherein the system (1) further comprises a discriminator module (3) that gets both the stream of events (21a) from the vision sensor (2) and information (52) about the heading and/or speed of the motion of the primary object (50) as inputs, and is configured to identify, from said stream of events (21a), based at least in part on said information (52), events (21b) that are likely to be caused by the motion of a secondary object (55), rather than by the motion of the primary object (50). Vision sensors (2) for use in the system (1). A corresponding computer program.

In one embodiment, a method includes sending a set of instructions to present, on a computing device, one or more available locations for a vehicle to pick-up or drop-off a user in an area. The one or more available locations are based on sensor data of the area that is captured by the vehicle. The method includes receiving a user selection to select a location associated with the area for the vehicle to pick-up or drop-off the user. The method includes adjusting a viability value of one or more locations to pick-up or drop-off the user. The viability value is adjusted based at least on the selected location. The method includes, based on the adjusted viability value of the one or more locations, determining a location from the one or more locations. The method includes instructing the vehicle to travel to the determined location.

A deep neural network(s) (DNN) may be used to detect objects from sensor data of a three dimensional (3D) environment. For example, a multi-view perception DNN may include multiple constituent DNNs or stages chained together that sequentially process different views of the 3D environment. An example DNN may include a first stage that performs class segmentation in a first view (e.g., perspective view) and a second stage that performs class segmentation and/or regresses instance geometry in a second view (e.g., top-down). The DNN outputs may be processed to generate 2D and/or 3D bounding boxes and class labels for detected objects in the 3D environment. As such, the techniques described herein may be used to detect and classify animate objects and/or parts of an environment, and these detections and classifications may be provided to an autonomous vehicle drive stack to enable safe planning and control of the autonomous vehicle.

In one embodiment, a method includes receiving sensor data corresponding to an environment external of a vehicle. The sensor data include data points. The method includes determining one or more subsets of the data points. The method includes comparing the one or more subsets of the data points to one or more predetermined data patterns. Each of the one or more predetermined data patterns corresponds to an object classification. The method includes computing a confidence score for each subset of data points of the one or more subsets of the data points as corresponding to each of the one or more predetermined data patterns based on the comparison. The method includes generating a classification for an object in the environment external of the vehicle based on the confidence score.