【作者】 曹瑜晗； 仝璐； 沈志昕； 杨柏桦； 吴礼；

【Author】 Yuhan Cao;Lu Tong;Zhixin Shen;Baihua Yang;Li Wu;School of Electronic Engineering and Optoelectronic Technology,Nanjing University of Science and Technology;Shanghai Radio Equipment Research Institute;

【机构】 南京理工大学电子工程与光电技术学院； 上海无线电设备研究所；

【摘要】 针对传统汽车乘员类型识别方法存在的准确度低、稳定性差等问题,本文提出了一种基于Faster R-CNN的毫米波雷达乘员识别算法。该算法在对毫米波调频连续波(Frequency Modulated Continuous Wave,FMCW)雷达车内乘员回波信号处理的基础上,提取了距离-角度热度图,构建了数据特征集,并设计了基于Faster R-CNN的识别算法,实现对车内乘员类型的识别。通过实验对算法性能进行评估,实验结果表明提出的识别算法对成员类型的识别准确率高达98%。更多还原

【Abstract】 Aiming at the problems of low accuracy and poor stability of traditional occupant recognition methods,this paper proposed an algorithm for occupant recognition of millimeter wave radar based on Faster R-CNN.After processing millimeter wave frequency modulated continuous wave(FMCW) radar echo signal of the occupant,the range-angle heatmap is extracted to construct a data feature set and the recognition algorithm based on Faster R-CNN is designed to realize occupant recognition.The performance of the algorithm is evaluated by experiments,and the experimental results show that the recognition accuracy of the proposed algorithm is up to98%.更多还原