Abstract：On-line handwritten signature verification is one kind of biometric identity recognition technology. Response rate of verification is a key factor for a handwritten signature verification system as well as recognition correct rate and noise robust property. An improved DTW algorithm is discussed to reduce the recognition time, which optimizes the conventional DTW algorithm's structure with the method of dynamic programming, and fastens the template matching speed in real-time handwriting identification. At the same time this paper puts forward a weighted distance calculation formula based on weighting to the stability of every stroke. The experiment has proved that this method is effectual. This method reduces the False Reject Rate(FRR) while guaranteeing high recognition rate. The matching distance is confined properly when the number of models increases. The experimental results show that the algorithm not only reduces the amount of calculation data, but also minimizes the duration of the whole processing. The advantage of the new algorithm is more obvious with the increment of models. Experiments show that this algorithm is highly efficient in computation, and generates nice result of stereo matching.