Ultrasound shear wave imaging is a rapidly emerging technique that noninvasively quantifies shear wave speed (SWS), which is directly linked to the stiffness of soft tissues. Targeting sources of estimation errors entailed in in vivo data, previous studies have proposed different methods, e.g. radon sum (RS) (Rouze et al. 2010), under the assumption of tissue homogeneity, which significantly limits its application in tissues with complex structure. In this study, we propose dynamic programming regularized wavelet transform (DPRWT) towards robust estimation of local SWS in inhomogeneous and anisotropic soft tissues.