Scalable Real-Time Flock Detection

Smart Cities services and applications can generate massive amounts of data to be processed in real-time, including the ones regarding moving objects. The ability to handle fast streams of data, especially detecting movements of groups of objects, is of paramount importance for the advance of such services. In this paper, we address the challenge of scalability for real-time detection trajectories of flocks, in the context of smart cities. The Bitmap Disk Filtering (BitDF) is an scalable algorithm capable of handling large-scale datasets. Performance evaluation results shows that, when compared to the current state-of-the-art algorithm, BitDF can reduce 99% of the processing time while not affecting accuracy. We prove the efficiency of BitDF by performing evaluation with both real and synthetic large datasets. BitDF is then very suitable to meet the requirements of real-time applications for a variety of trajectory-based applications.