A handheld swept optical coherence tomography system capable of rapidly imaging subcutaneous human microvascular is established. Three algorithms are used to reconstruct the blood flow distribution images: power intensity differential (PID) without logarithmic compensation, logarithmic-compensated power intensity differential (LCPID), and logarithmic-compensated power intensity differential with motion threshold (MTPID). Then, microvascular en-face images obtained by three imaging algorithm are compared. The results show that the LCPID algorithm can reveal deeper blood flow information, and the MTPID algorithm can get more details of the blood flow distribution and high image resolution. So, the superiority of MTPID algorithm has significance for the application of OCT system in biomedical optics.