Bike A Step Towards Safer Green Transportation S Smaldone C Tonde V K Ananthanarayanan A Elgammal L Iftode Summarized by Yuki Outline Analysis Design Implementation ID: 173706
Download Presentation The PPT/PDF document "The Cyber-Physical" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
The Cyber-Physical Bike:A Step Towards Safer Green Transportation
S. SmaldoneC. TondeV. K. AnanthanarayananA. ElgammalL. IftodeSummarized by YukiSlide2
OutlineAnalysisDesign
ImplementationEvaluationConclusionSlide3
AnalysisProblemA
biker must spend cognitive and physical ability to periodically scan for a rear-approaching vehicle. As a result, the biker can not maintain continual awareness for the forward situation.SolutionThe Cyber-Physical bicycle system continuously monitors the environment behind the biker, automatically detects rear-approaching vehicles, and alerts the biker prior to the approach.Slide4
DesignRoadway segmentation analysis using o
ptical flowReduces computational loadSlide5
DesignVehicle trackingDetermines
if each vehicle passes the biker in either a safe or unsafe mannerSlide6
ImplementationDevelopment environment
C/C++NVIDIA CUDA library (v2.3)Open source optical flow library[1] Bike computer - HP Mini 311 notebookIntel Atom N280 1.67 GHz CPU3 GB RAMNVIDIA ION GPU (16 CUDA cores and 256 MB memory)80 GB SSD hard disk3.26 pounds / 1.48 kilograms (the
worst case with peripherals)
1.
WERLBERGER, M., TROBIN, W., POCK, T., WEDEL, A., CREMERS, D., AND BISCHOF, H. Anisotropic Huber-L1 optical flow. In BMVC’09 (London, UK, 2009).Slide7
Implementation (cont.)Rear-facing video camera - Sony DCR-SX40
Over 3 hours of video recordings of real-world roadway cycling traces (repeatability of experiment)Every interaction between the biker and a vehicle (approaching and departing) are manually annotated using the timestampsOrdinary road bicycle - Trek FX7.5Slide8
EvaluationHow accurate are video-based techniques in
detecting rear-approaching vehicles?Accuracy = TP / (TP + FP + FN) = 19 / (19 + 6 + 1) = 73.1%A vehicle hidden by another biker
A
sudden
vibration caused by uneven roadwaySlide9
Evaluation (cont.)Can detection be performed timely?
Alerting to occur an average of 3.5 seconds prior to a vehicle encounter at reduced frame rates (3 FPS)3.5 seconds is 92% of the potential time (the difference between the first appearance of the vehicle and the time it passes the cyclist)Slide10
ConclusionThe Cyber-Physical bicycle is a system that augments normal bicycles
with video processing capabilities for automated rear-approaching vehicle detection.This system directly improves the safety of bikers by reducing their cognitive overheads of continuously probing for rear-approaching vehicles.The prototype can perform rear-vehicle detection with good accuracy at full frame rates (30 FPS), and can operate in real-time at reduced frame rates (3 FPS).