Coronavirus disease (COVID-19), which was caused by SARS-CoV-2, has become a global public health concern. A great proportion of the world needs to be vaccinated in order to stop the rapid spread of the disease. In addition to prioritising vulnerable sections of the population to receive the vaccine, an ideal degree-based vaccination strategy uses fine-grained contact networks to prioritise vaccine recipients. This strategy is costly and impractical due to the enormous amount of specific contact information needed. It also does not capture indirect famine or aerosol-based transmission. We recently proposed a new vaccination strategy called Individual's Movement-based Vaccination (IMV), which takes into account both direct and indirect transmission and is based on the types of places people visit. IMV was shown to be cost-efficient in the case of influenza-like diseases. This paper studies the application of IMV to COVID-19 using its documented transmission parameters. We conduct large scale computer simulations based on a city-wide empirical mobility dataset to evaluate the performance and practicability of the strategy. Results show that the proposed strategy achieves nearly five times the efficiency of random vaccination and performs comparably to the degree-based strategy, while significantly reducing the data collection requirements.