Abstract: A method of detecting lanes includes the steps: capturing (S1) a camera image (K) of a vehicle environment by a camera device (2) of a vehicle (5); determining (S2) feature points (P1 to P15) in the camera image (K), which feature points correspond to regions of possible lane boundaries (M1, M2); generating (S3) image portions of the captured camera image (K) respectively around the feature points (P1 to P15); analyzing (S4) the image portions using a neural network to classify the feature points (P1 to P15); and determining (S5) lanes in the vehicle environment taking account of the classified feature points (P1 to P15).

Abstract: A system for generating at least a second trajectory for a first route section of a road comprises a first interface for receiving first data representing at least a first trajectory. The first data were recorded during travel on the first route section by at least one vehicle controlled by a human driver. The first interface is additionally configured to receive second data representing ambient conditions at the time of recording of the first trajectory, and to receive third data representing vehicle-related features present during the recording of the first trajectory. The system additionally comprises a first data processing module that performs clustering of multiple first trajectories based on associated second data and/or third data, a database for retrievably storing the results of the clustering, and a second interface for receiving a request for the transmission of a second trajectory and for corresponding transmission of the requested second trajectory.

Abstract: The invention relates to a method for detecting lanes (F), having the steps of: capturing (S1) a camera image (K) of a vehicle environment by means of a camera device (2) of a vehicle (5); determining (S2) feature points (P1 to P15) in the captured camera image (K), which correspond to regions of possible lane boundaries (M1, M2); generating (S3) image portions of the captured camera image (K) around each feature point (P1 to P15); analyzing (S4) the image portions using a neural network in order to classify the feature points (P1 to P15); and determining (S5) lanes in the vehicle environment taking account of the classified feature points (P1 to P15).

Abstract: A method for estimating traffic lanes uses as input data the position and direction of feature vectors, which are measured from multiple different sensors independently of one another. A feature vector is formed by a position in the coordinate system of the ego-vehicle, which position describes a point on the border of a traffic lane, and a direction or an angle which indicates the direction in which the border of the traffic lane runs at this position. Further input data are variables which represent the quality of the measurement of the positional and the directional accuracy and the probability of the existence of a feature vector. The input data are accumulated chronologically together. The geometry of traffic lanes is estimated from the accumulated input data, taking into account the quality of the measurement. The estimated geometry of traffic lanes is output.

Abstract: A method for automatically establishing extrinsic parameters of a camera of a vehicle includes repeatedly establishing the extrinsic parameters of the camera during a measurement run, wherein sensor data are generated by a sensor each time the extrinsic parameters of the camera are established. The established parameters and the associated sensor data respectively are stored in a database. A convolutional neural network is trained based on the extrinsic parameters stored in the database and based on the sensor data. The extrinsic parameters of the camera are subsequently determined online by utilizing the trained convolutional neural network.

Abstract: A system for generating at least a second trajectory for a first route section of a road comprises a first interface for receiving first data representing at least a first trajectory. The first data were recorded during travel on the first route section by at least one vehicle controlled by a human driver. The first interface is additionally configured to receive second data representing ambient conditions at the time of recording of the first trajectory, and to receive third data representing vehicle-related features present during the recording of the first trajectory. The system additionally comprises a first data processing module that performs clustering of multiple first trajectories based on associated second data and/or third data, a database for retrievably storing the results of the clustering, and a second interface for receiving a request for the transmission of a second trajectory and for corresponding transmission of the requested second trajectory.

Abstract: A method for estimating traffic lanes uses as input data the position and direction of feature vectors, which are measured from multiple different sensors independently of one another. A feature vector is formed by a position in the coordinate system of the ego-vehicle, which position describes a point on the border of a traffic lane, and a direction or an angle which indicates the direction in which the border of the traffic lane runs at this position. Further input data are variables which represent the quality of the measurement of the positional and the directional accuracy and the probability of the existence of a feature vector. The input data are accumulated chronologically together. The geometry of traffic lanes is estimated from the accumulated input data, taking into account the quality of the measurement. The estimated geometry of traffic lanes is output.

Abstract: A method for automatically establishing extrinsic parameters of a camera of a vehicle includes repeatedly establishing the extrinsic parameters of the camera during a measurement run, wherein sensor data are generated by a sensor each time the extrinsic parameters of the camera are established. The established parameters and the associated sensor data respectively are stored in a database. A convolutional neural network is trained based on the extrinsic parameters stored in the database and based on the sensor data. The extrinsic parameters of the camera are subsequently determined online by utilizing the trained convolutional neural network.