The rapid development of cognitive neuroscience makes objective determination of various physiological parameters possible. Functional near-infrared spectroscopy (fNIRS) is an emerging brain imaging tool that can detect human skin tissue hemodynamic indices including oxygenated hemoglobin (HbO), deoxygenated hemoglobin (Hb) and total hemoglobin (tHb). Electrocardiograph (ECG) and respiration wave (RSP) are also two important physiological parameter determination methods. In order to obtain multiple physiological parameters using fNIRS only, three algorithms, including time-domain waveform characteristic analysis, frequency-domain band-pass filtering and wavelet decomposition and reconstruction, are used to calculate heart rate (HR) and breath rate (BR) based on the HbO data collected by fNIRS. The calculated HR and BR results are compared with the real HR (77.0199) surveyed by ECG and BR (22.9153) surveyed by RSP. The results show that the three methods can all extract HR from fNIRS effectively, wherein the band-pass filtering can extract the most accurate HR (76.8807) with the deviation of -0.1392. The BR (21.7039) with the deviation of -1.2114 is also calculated by the same algorithm. Extracting HR and BR features using fNIRS signal is realized.