Contactless detection of human beings via extracting vital sign features (VSF) is a perfect technology by employing an ultra-wideband radar. Only using Fourier transform, it is a challenging task to extract VSF in a complex environment, which can cause a lower signal to noise ratio (SNR) and significant errors due to the harmonics. This paper proposes an improved signal processing algorithm for VSF extraction via analyzing the skewness and standard deviation of the collected impulses. The discrete windowed Fourier transform technique is used to estimate the time of arrival of the pulses. The frequency of human breathing movements is obtained using an accumulation scheme in frequency domain, which can better cancel out the harmonics. The capabilities of removing clutters and improving SNR are validated compared with several well-known methods experimentally.

中文翻译：

---

一种用于 VSF 提取的改进信号处理算法

通过提取生命体征特征（VSF）对人体进行非接触式检测是一种采用超宽带雷达的完美技术。仅使用傅立叶变换，在复杂环境中提取 VSF 是一项具有挑战性的任务，这会导致较低的信噪比 (SNR) 和由于谐波引起的显着误差。本文通过分析采集到的脉冲的偏度和标准偏差，提出了一种用于VSF提取的改进信号处理算法。离散加窗傅立叶变换技术用于估计脉冲的到达时间。人体呼吸运动的频率是通过频域累加方案获得的，可以更好地抵消谐波。