Video upscaling is a common task performing the adaption of the actual video resolution to the screen display area. The extensively used interpolation methods are easily implemented but usually result in some artifacts, such as blurring and aliasing effects. This paper proposes to combine sharpening with interpolation to reduce the blurring effects of enlarged videos, in which videos are firstly preprocessed with sharpening to enhance profiles so as to compensate the loss of high-frequency components during interpolation. To tune sharpness degree in sharpening procedure, the blurriness metric considering image gradients as well as pixel intensities is particularly suggested and the generation rule for sharpening strength is accordingly regulated. In practice, three representative interpolation methods, namely, Bilinear, Bicubic and Lanczos, are experimentally incorporated into our framework to carry out validation. Results with respect to both original and decoded videos show the significant objective and subjective improvements against the respective regular interpolation.