Blood flow measurement provide necessary information for the diagnosis of diseases causing blood flow disorders. Ultrasound scatter de-correlation technology is a promising technique for the velocity measurement of blood flow. To achieve more accurate blood flow velocity measurement with scatter de-correlation, a new signal processing method is proposed. The proposed method is based on dynamic time warping of two consecutive radio frequency signals, and tested on an experimental platform based on high-frequency intravascular ultrasound data acquisition system (50 MHz). Experimental results show that, compared with the traditional windowed cross-correlation method, the proposed method can significantly improve the accuracy of the time shift calculation, leading to more accurate blood flow velocity measurement. The maximum error of the proposed method is less than 5% at 10 speeds in the range of 4.72 to 25.95 cm/s. With the usage of the proposed method, the average error and covariance are reduced by 68% and 57%, respectively, compared with windowed cross-correlation.