A Study of Robot Ground Classification Based on SMOTE Oversampling Technique and XGBoost

Accurate identification of ground types during robot travelling is crucial for improving the robots navigation stability and decision-making ability. To this end, this paper proposes a robot ground type classification method based on SMOTE oversampling technique and XGBoost, specifically, we firstly balanced the unbalanced dataset by SMOTE technique, and thereafter fed the processed data into XGBoost model for classification. We conducted extensive comparative experiments comparing common machine learning algorithms and found that the 91% accuracy of our algorithm achieved the best results, which proves the advancement and superiority of our proposed method and provides effective technical support for autonomous robot navigation and environment sensing.