Performance Optimization of 3D Human Body Reconstruction Technology Based on mmWave Radar

Wireless Sensing Technology utilizes radio waves, microwaves, infrared and other wireless signals to sense and detect the environment. Millimeter wave(mmWave) radar is widely applied in the field of wireless perception due to its high stability and adaptability. Therefore, the 3D body reconstruction technology based on mmWave radar can adapt to extreme environments better than other sensors. However, the performance of existing 3D human body reconstruction methods with mmWave radar is not particularly ideal, especially the accuracy of the limbs is not good. In this paper, we improve the 3D human body perception and reconstruction performance of mmWave radar by improving the loss strategy and using multi-frame joint correction. By using loss strategies that have better effects, we improve the accuracy of human reconstruction. And by detecting abnormal frames through differences between adjacent frames and utilizing the information between frames for correction, we significantly improved the handling of extreme anomalies. The effectiveness of our method is verified by the comparative analysis of many metrics and the intuitive feeling of the visualization effects.