VIRTUe: Video surveillance for rail-road traffic safety at unmanned level crossings; (Incorporating Indian scenario)

Railway Transportation is one of the most discussed topics in terms of safety and security. This originate need to install safety elements to avoid accidents. In this paper, we scrutinize various level crossing conditions and trying to give an effective solution to the current Indian rail-road scena...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:2017 IEEE Region 10 Symposium (TENSYMP) s. 1 - 4
Hlavní autori: Gajbhiye, Pranjali, Naveen, C., Satpute, Vishal R.
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 01.07.2017
Predmet:
Accidents
Algorithm design and analysis
Danger area
five frame background subtraction
five frame differencing
Level crossing
multiple object tracking
Object detection
Object tracking
Rail transportation
Safety
Surveillance
variance method
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Railway Transportation is one of the most discussed topics in terms of safety and security. This originate need to install safety elements to avoid accidents. In this paper, we scrutinize various level crossing conditions and trying to give an effective solution to the current Indian rail-road scenarios using visual surveillance. Here the smart visual surveillance system starts with separating, detecting, and tracking of moving objects in the level crossing (Danger area) region using the proposed variance-based method. This new method is based on five frame background subtraction, five frame differencing and variance calculation for object detection and tracking. In the variance based method, the variance of rows and columns in video frames are calculated where the variation in image intensity pixels gives the position of moving object and it is used for locating and tracking the objects in the video. The algorithm is tested on various hazardous conditions over the level crossings such as vehicle crossing over the region of interest, pedestrian crossing the ROI, and stopped vehicle over the ROI. On the basis of comparison between proposed variance based algorithm and mean shift method, variance based method is more robust to environmental noise and accurately detecting an object in danger area with minimum computational timings.
DOI:10.1109/TENCONSpring.2017.8070015