Intelligent Abnormal Situation Awareness Platform (i-ASAP)

Details

• Creators:
  Qian, Yu ; Wang, Yi ; Tang, Youzhi ; Song, Ge
• Corporate Creators:
  University of South Carolina
• Corporate Contributors:
  United States. Department of Transportation. Federal Railroad Administration
• Subject/TRT Terms:
  • Grade crossing protection systems
  • Monitoring
  • Behavior
  • Pedestrian safety
  • Computer vision
  • Artificial intelligence
• Publication/ Report Number:
  DOT/FRA/ORD-26/24
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States
• Edition:
  08/2020-08/2023
• Contracting Officer:
  Blue, Shala
• Corporate Publisher:
  United States. Department of Transportation. Federal Railroad Administration
• Abstract:
  This report introduces the Intelligent Abnormal Situation Awareness Platform (i-ASAP), a cutting-edge system developed to mitigate trespassing incidents at railroad crossings. The platform employs advanced computer vision and artificial intelligence technologies to monitor and identify trespassers and obstructions in real-time. Using AlphaPose and OpenPose models within a deep learning anomaly detection system, i-ASAP tracks pedestrian behavior to pinpoint anomalies. A newly developed weakly supervised foreground segmentation network extends the platform's detection capabilities to static and moving objects. Validation through custom dataset testing and real-time field trials demonstrate i-ASAP's effectiveness, setting the stage for potential deployment on edge computing devices.
• Format:
  PDF
• Funding:
  693JJ620C000021
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/93429/dot_93429_DS1.pdf
• File Type:
  [PDF - 5.42 MB]
• Collection(s):
  Federal Railroad Administration
• Main Document Checksum:
  urn:sha-512:b219006e0f7882037c9f0f1beb551c837a01d1107fef5a52ee3572aa2a1bb1383b341b69a3002bb70b51b8aaf856b5ee4f9b499cac57d45bb1350631b435ae75