The objective video quality metrics have been researched for years and many methods have been proposed. As a main feature of the Human Visual System (HVS), visual attention (or Region of Interest-ROI) will influence viewer's subjective feeling since artifacts on a ROI is much more annoying than those appearing on an inconspicuous area. However, little study has been taken on identifying how and to what extent ROI will influence video quality measurements. In this paper, we propose a fully automatic region of interest weighted pooling strategy considering the influence of visual attention, which is then evaluated on VQEG Phase I FR-TV test dataset. Apparent and coherent performance improvement is achieved by applying the proposed pooling strategy on PSNR and SSIM, together with a highly reduction in computation complexity.