In time series mining, the Dynamic Time Warping (DTW) distance is a commonly and widely used similarity measure. Since the computational complexity of the DTW distance is quadratic, various kinds of warping constraints, lower bounds and abstractions have been developed to speed up time series mining under DTW distance.