Research on Target Quantity and Distance Identification Method for Through-Wall Radar Based on CNN

The deep learning technology is combined with the through wall radar imaging method, and the research is carried out through the combination of simulation and experiment. The specific contents include:gprMax based through wall radar sim-ulation. gprMax software is used to establish the detection model of through wall radar and realize the simulation. The simulation data are collected and the appropriate method is selected for clutter filtering, which lays the foundation for the radar data processing obtained from the experiment. Through wall radar experiment: use the life detector radar of the laboratory to do the actual scene experiment, and obtain the experimental data of different people at different distances in the detection area, which will be converted into pictures for later processing. Establish neural network model. The radar echo image obtained from the experiment is cut and processed and sent to the neural network for training, and the number and distance of targets in the image are correctly predicted.