A control method of a vehicle may include generating navigation information based on destination information and current location information; determining whether a lane to be driven is a merge lane based on the generated navigation information, map information, and the current location information; recognizing a lane in an image acquired by an imaging device; recognizing a driving first lane based on the recognized location information of the lane; dividing a certain area including the merge lane into an entry section, a merge section, and a stabilization section when the first lane converges with a second lane; generating a driving route by performing curve fitting for route points in the entry section and the merge section; and controlling autonomous driving based on the generated driving route.

A construction machine safety management system includes an information generating unit that generates information representing a relationship between a detection result and predetermined related information, based on the detection result obtained by a detecting unit that detects entry of an obstacle into a predetermined area around a construction machine and the predetermined related information corresponding to the detection result.

The roadside object detection device, roadside object detection method, and roadside object detection system according to the present invention make it possible to accurately detect roadside objects by: determining a first feature value regarding positional relationship between a host vehicle and a roadside object candidate and a second feature value regarding a height of the roadside object candidate, based on external environment information acquired by an external environment recognition unit; determining a height-related threshold for identifying the roadside object candidate as a roadside object, based on the first feature value; and identifying the roadside object candidate as a roadside object when the second feature value exceeds the threshold.

A stereo camera including a pair of lens units, an imaging sensor which obtains a pair of images via the pair of lens units, a distance calculation unit which calculates distances of a plurality of positions within a detection area, which is an area where the pair of images are overlapped, based on the pair of images, and a distance correction unit which applies correction to the calculated distances of the plurality of positions, wherein the correction corresponds to positions in an arrangement direction of the pair of lens units, the correction is ne by which the closer the position comes close to an end portion in the arrangement direction within the detection area, the larger a reduction amount of the calculated distances by the correction becomes.

A method for depth estimation performed by a depth estimation system of an autonomous agent includes determining a first pose of a sensor based on a first image captured by the sensor and a second image captured by the sensor. The method also includes determining a first depth of the first image and a second depth of the second image. The method further includes generating a warped depth image based on at least the first depth and the first pose. The method still further includes determining a second pose based on the warped depth image and the second depth image. The method also includes updating the first pose based on the second pose and updating a first warped image based on the updated first pose.

The present technology relates to an imaging device, a signal processing device, and a vehicle control system enabling proper driving support when a vehicle enters a road from the outside of the road. An imaging device includes: an imaging unit that captures an image of a front of a vehicle; and an object detection unit that performs object detection processing on the basis of the image captured by the imaging unit, wherein the object detection unit changes an object detection method on the basis of a positional relationship between the vehicle and a road where the vehicle enters from an outside. The present technology can be applied to an imaging device installed in, for example, various vehicles that perform driving support.

An image processing apparatus includes an edge detector and a sidewall information calculator. The edge detector detects outlines extending from lower left toward upper right with respect to objects located on left side of a vehicle, and detects outlines extending from lower right toward upper left with respect to objects located on right side of the vehicle, on the basis of a luminance value of an image out of a pair of images captured by a stereo camera. The sidewall information calculator recognizes, on the basis of three-dimensional positional information of the objects having the outlines detected by the edge detector, whichever of the objects having the outlines detected by the edge detector is located at a predetermined height or greater from a road surface and has a predetermined length or greater in a front-rear direction of the vehicle, as a sidewall.

The present invention discloses a grid map obstacle detection method fusing probability and height information, and belongs to the field of image processing and computer vision. A high-performance computing platform is constructed by using a GPU, and a high-performance solving algorithm is constructed to obtain obstacle information in a map. The system is easy to construct, the program is simple, and is easy to implement. The positions of obstacles are acquired in a multi-layer grid map by fusing probability and height information, so the robustness is high and the precision is high.

In various examples, live perception from wide-view sensors may be leveraged to detect features in an environment of a vehicle. Sensor data generated by the sensors may be adjusted to represent a virtual field of view different from an actual field of view of the sensor, and the sensor data—with or without virtual adjustment—may be applied to a stereographic projection algorithm to generate a projected image. The projected image may then be applied to a machine learning model—such as a deep neural network (DNN)—to detect and/or classify features or objects represented therein. In some examples, the machine learning model may be pre-trained on training sensor data generated by a sensor having a field of view less than the wide-view sensor such that the virtual adjustment and/or projection algorithm may update the sensor data to be suitable for accurate processing by the pre-trained machine learning model.

A system and method for using static objects of an area detected during an offline map generation process to aid the online perception algorithm of an autonomous driving vehicle (ADV) is disclosed. 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 predefined static objects may be added back to the output of the perception algorithm to complete the perception output.