To control and contain the outbreaks of emerging infectious diseases such as COVID-19, it is important to know how easy and fast they transmit among people. To explore the essential information of the novel infectious agents, people always confront an inverse problem: using (partially) observed numbers of infected people by time and region to dig up the underlying characteristics of unknown infectious agents. Epidemics armed with advanced statistical inference and mathematical theory has been developed to help reconstruct transmission dynamics processes and to estimate key features of infectious diseases. In this study we use COVID-19 outbreak in Shaanxi province as an example to illustrate how the infectious disease dynamics method can be used to help build the transmission process and to estimate the transmissibility of COVID-19. Three transmission dynamics models were proposed for this. By separating continuous importation from local transmission and treating imported cases as the source rather than results of local transmission, the basic reproduction number of COVID-19 in Shaanxi province was estimated in the range from 0.46 to 0.61, well below the critical value of 1.0. This indicates that COVID-19 cannot self-sustain in Shaanxi province and reflects the timely and strong control measures taken in Shaanxi province.