eduzhai > Applied Sciences > Engineering >

VidCEP Complex Event Processing Framework to Detect Spatiotemporal Patterns in Video Streams

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: Video data is highly expressive and has traditionally been very difficult fora machine to interpret. Querying event patterns from video streams ischallenging due to its unstructured representation. Middleware systems such asComplex Event Processing (CEP) mine patterns from data streams and sendnotifications to users in a timely fashion. Current CEP systems have inherentlimitations to query video streams due to their unstructured data model andlack of expressive query language. In this work, we focus on a CEP frameworkwhere users can define high-level expressive queries over videos to detect arange of spatiotemporal event patterns. In this context, we propose: i) VidCEP,an in-memory, on the fly, near real-time complex event matching framework forvideo streams. The system uses a graph-based event representation for videostreams which enables the detection of high-level semantic concepts from videousing cascades of Deep Neural Network models, ii) a Video Event Query language(VEQL) to express high-level user queries for video streams in CEP, iii) acomplex event matcher to detect spatiotemporal video event patterns by matchingexpressive user queries over video data. The proposed approach detectsspatiotemporal video event patterns with an F-score ranging from 0.66 to 0.89.VidCEP maintains near real-time performance with an average throughput of 70frames per second for 5 parallel videos with sub-second matching latency.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...