Research on Radar Target Classification Algorithm Based on Multi-feature Fusion and Deep Learning
Chengxin Zhang, Ao Wang, Yijin Zhang, Weibin Zhang · 2022 34th Chinese Control and Decision Conference (CCDC) · 2022
Whether radar can correctly classify objects plays an important role in the field of autonomous driving. This paper proposes an algorithm of radar target classification method based on multi-feature fusion and deep learning. Then we define a feature score based on the methods of random forest and correlation analysis for feature dimensionality reduction. After that, the sliding window is adopted for data enhancement. Then the data is trained by using the LSTM (Long Short-Term Memory) and CNN (Convolutional Neural Network) for target classification. Finally, we use the average accuracy of the two models to fuse the classification results to increase the final accuracy. The test results prove the effectiveness of the proposed algorithm.