Abstract: Reverse time migration (RTM) based on the elastic wave equation in Tilted Transversely Isotropic (TTI) medium has significant advantages in imaging of middle and deep complicated geological structures. However, TTI RTM, being rooted in a two-way wave equation, inevitably generates substantial low-wavenumber noise when applying cross-correlation imaging conditions. Although the cross-correlation imaging conditions based on the Poynting vector wavefield decomposition can suppress the low-wavenumber noise to a certain extent, it is difficult to accurately indicate the propagation direction of the wave field, and it is easy to appear instability in the calculation process. In this study, we applied the wavefield decomposition technology based on optical flow vector to the TTI medium elastic wave reverse time migration. Based on the optical flow vector, the source wavefields and receiver wavefields could be decomposed clearly. With the cross-correlation imaging conditions and the weighted imaging strategy, the high precision imaging section could be obtained. The processing of the real seismic data from the Yellow Sea area showed that compared to the Poynting vector, the optical flow vector could more accurately and stably achieve the decomposition of elastic waves in TTI medium, based on which the cross-correlation weighted processing of the decomposed section could more effectively suppress low wavenumber noise in the imaging section and improve imaging quality, show wide application prospects of this new technology.