The initial Sequential Monte Carlo cardinality balanced multi-target multi-Bernoulli (SMC-CBMeMBer) filter assumes that the target birth intensity is known as a priori, but in fact the new target may appear anywhere in the detected area. And SMC-CBMeMBer does not extract track label information and cannot distinguish the target. Therefore, this paper proposes a labeled SMC-CBMeMBer filter with adaptive track initiation (SMC-CBMeMBer-D-L-ATI). Firstly, the filter introduces the track label information in the prediction process, and selects the measurement that may correspond to the newborn target, obtaining the initial state of the newborn target according to the converted measurement. Then, the label information in the prediction process is inherited during the update process, and the target states are updated sequentially through the converted position measurement and Doppler measurement. The simulation proves that the proposed filter can effectively perform adaptive track initiation, with good tracking performance, and can provide the target track information.