In the emerging big data era, in addition to explicit publication of users' locations on geo-social networks, positioning system embedded in mobile phones implicitly records users' locations. Although such implicitly collected spatiotemporal data play an important role in a wide range of applications such as disease outbreak control and route recommendation for life science or smart city, they cause new serious privacy issues when cross-referencing with the explicitly published data from users. Existing location privacy preservation techniques fail to preserve the proposed implicit privacy because they ignore the cross-reference between implicitly and explicitly spatiotemporal data. To tackle this issue, this work for the first time investigates and defines the implicit privacy and proposes the discover and eliminate framework. In particular, this paper proposes prefix filtering based nest loop algorithm and frequent moving object based algorithm to generate dummy data to preserve the proposed implicit privacy. Further, it constructs an improved reverse a priori algorithm and graph based dummy data generation algorithm respectively to make the solution more practical. The results of extensive experiments on real world datasets demonstrate the effectiveness and efficiency of the proposed methods.