Efficient Real-time Query of Multimedia Data Based on Local Sensitive Hash

In recent years, monitoring of modern metropolitan events and civil infrastructure has been identified as an important part of the management of public services. Through rapid information from citizens, intelligent equipment for collecting multimedia information can pass from a terminal initialization, application, examination, provide useful information to deduce such emergency situations. To this end, in this paper, we proposed a realtime query framework in IoT sensing data streams for efficient realtime query of multimedia with hash method. More specifically, the proposed method is featured by: (i) the Batch Orthogonal Locality Sensitive Hashing is adopted to deal with the image and video data with hash code. (ii) a comprehensive flow processing is designed for realtime data query based on storm and kafka. Finally, a series of experiments are presented to evaluate the efficiency and satisfactory performance of the proposed methods.