Abstract: In approximately adiabatic shallow water environment, comparing to traditional Matched Field Processing localization method, target localization using virtual receiver method can reduce dependence on environment parameters and acoustic field model. Also, it can avoid large computation of copy field. Virtual receiver ranging method locates the target by estimating the slope of the interference striations. Incomplete receiver array disturbs the interference structure and degrades the localization performance. A Multi-scale Linear Filter(MLF) based on image Hessian Matrix eigenvalue analysis is introduced. Numerical simulation and trial data processing demonstrate that MLF can be applied in localizing the target source with incomplete receiver array using virtual receiver method as it can effectively strengthen and extract the information in the interference striations.