Method of Classifying a Road Surface Object, Method of Training an Artificial Neural Network, and Method of Operating a Driver Warning Function or an Automated Driving Function

A computer-implemented method of training an artificial neural network to classify road surface objects, the method comprising:

• providing a plurality of data points, each data point specifying a first variable and a second variable, wherein for each data point
• the first variable characterizes a horizontal motion exhibited by a vehicle when driving over the road surface object;
• and
• the second variable characterizes a vertical motion exhibited by said vehicle when driving over the road surface object;
• providing ground truth data indicating a respective relevance of a plurality of road surface objects for a driver warning function or an automated driving function;
• clustering the data points into a plurality of sets of data points, such that each data point of a same set of data points is attributable to a same road surface object;
• matching the ground truth data with the sets of data points so as to obtain a plurality of sets of data points, each set of data points being attributable to a respective road surface object and to a relevance of the road surface object for a driver warning function or an automated driving function;
• for each set of data points, generating an image data file depicting the data points in a form of a scatter plot;
• and
• training an artificial neural network to classify road surface objects by using the image data files and the corresponding ground truth information regarding the relevance of the respective road surface objects as training data.