为准确实时地检测出运动过程中人体的心率值, 改进传统的心率计算方法是非常必要的。文中介绍了基于频域计算的离散Fourier 变换(DFT ) 算法—— 改进的戈泽尔算法, 实现了时域信号的频域分析。它克服了运动过程中所带进的干扰、减少了必须用于DFT 算法计算所需的脉搏包络的数据长度和DFT 所需的时间。文中还介绍了使用近红外光传感器检测脉搏波的方法, 以及算法在8 bit 的EPSON 单片机上的实现, 并研制出了原型样机用于实际的检测。结果表明系统在实时脉搏检测上得到了比传统脉搏检测方法更 好的效果。 关键词: 心率; 戈泽尔算法; 离散傅立叶变换; 近红外光 Abstract: The tradit ional pulse detection method must be imp roved to detect the heart2rate of a moving body in real time. A frequency2based discrete Fourier transfo rm s (DFT) algo rithm, an imp roved Gerzel algo rithm, is introduced to analyze the signals in the frequency domain. The method can overcome disturbances and decrease the length of the pulse packet that must be given to the DFT algo rithm which reduces the calculation time. The pulse wave can be detected using a near infrared senso r and analyzed using by an 82bit EPSON m icrop rocesso r. The system has been used to detect the heart2rate at a moving body in real time w ith better results than traditionalmethods. Key words: heart2rate; gerzel algo rithm; DFT; near infrared