Abstarct: Structural interpretation on seismic images with low resolution is always accompanied by high uncertainty in positioning reflectors and focusing structures in their correct places. This uncertainty results in inaccurate geological interpretation and misunderstanding in subsurface geological model evaluation. Thus, reducing geological interpretation uncertainties is among advanced researches in seismic imaging. The main source of introducing uncertainty in seismic imaging comes from structural complexity of wave propagation media and lateral changes in velocity. Accurate positioning of reflectors and their dips strongly depends on the accuracy and precision of the used velocity model through seismic imaging. Interpretation studies have shown that even slight complexity in geological structure and mild lateral velocity change will result in large positioning and focusing error. Therefore it is required to apply a robust velocity model updating scheme in seismic imaging procedure. Among common methods for velocity analysis, the semblance analysis is known as the most common coherence measure with poor resolution due to its low sensitivity to velocity changes. However, its resolution could be increased by adding a special term to its semblance equation to increase its sensitivity to velocity changes. In this study, the weighted semblance method was selected for improving resolution of the velocity spectrum. It was used for weighting the common semblance method for time migration velocity analysis. To increase resolution of time migration velocity spectra that was generated by semblance method, we added weighting function baxsed on first to second singular values ratio of time analysis window to the semblance equation. This weighting function enhances resolution of migration velocity spectra in velocity direction. The proposed strategy was applied on a marine data to investigate its efficiency in reducing structural uncertainty and positioning error. Results in the first steps have shown that the velocity resolution was increased in velocity spectra. Obtained velocity model then was used for time migration. Subsequently, by calculating uncertainty in velocity model estimation and structural sensitivity of time-migrated image on velocity model, positioning error of reflectors and focusing uncertainty of geological structures were defined. Results of application the proposed method in comparison with results of common semblance method reveals that positioning errors in migrated seismic images were decreased due to increasing in resolution of time migration velocity spectra.