Road condition recognition in self-driving cars based on classification and regression tree

With the continuous development of technology, the realization of self-driving cars is getting closer and closer to our life. Obviously, road condition recognition is un-doubtedly the premise and component of this technology. In this paper, a new road condi-tion recognition method based on Classification And Regression Tree (CART) technology is realized. First, we process the original Oxbotica data to get a data set consisting of nine sensors information that can describe the vehicle’s driving. We then generate a CART model using the training data set. Finally, we test the model and calculate some evaluation indicators to evaluate our model. The results show that our method performs well for the identification of six types of road conditions.