Call For Papers
Contact Us

  Accident Detection in Live Surveillance  
  Authors : Shrey Gupta; Vandana Choudhary
  Cite as:


With the increase in number of vehicles in the country vehicle detection is an important in road traffic management system. Different traffic accident causes such as vehicle overspeeding, wrong way driving, collision and accident can be detected by CCTV installed on roads. The results obtained from traffic parameters can be applied for vehicle tracking, vehicle classification, parking area monitoring, road traffic monitoring and management etc. The main objective of this project is to decrease the deaths caused by accident occurring because over speeding, wrong war driving by ensuring public safety and also a building a better system for managing the traffic on the roads. The aim of this paper is to develop a system that can detect the vehicle accident which are caused by overspeeding, wrong way driving and collision detection on city roads. A prototype system is developed and tested.


Published In : IJCSN Journal Volume 8, Issue 3

Date of Publication : June 2019

Pages : 222-228

Figures :12

Tables : --


Shrey Gupta : IT, Maharaja Agrasen Institute of Technology, Rohini New Delhi, Delhi, India.

Vandana Choudhary : IT, Maharaja Agrasen Institute of Technology, Rohini New Delhi, Delhi, India.


Deep Learning, Heatmap, OpenCV, ROI, SSD Model, TensorFlow, RMS

This system is capable to detect speed of the car, collision prevention, wrong way direction and accident detection together. The system detects different causes of accidents and send alert to the authorities. It detects the speed of the car by capturing the image detected by the camera. The main aim of this project is to decrease the chances of loss of lives in accident occurring because over speeding, wrong way detection and collision detection and hence improve public safety and also a better system for the managing the traffic on the roads. The system is cost-effective, scalable, fast, at a distance measuring system that can be easily housed in present live surveillance. Apart from this, it also gives an opportunity for transportation engineers and decision makers to plan building of road and can be helpful both in case of personal as well as business purpose, to improves safety and security of the person on road.


[1] http://airccse.org/journal/jcsit/6214ijcsit16.pdf [2] https://www.tandfonline.com/doi/pdf/10.1080/164 84142.2004.9637965?needAccess=true [3] https://www.researchgate.net/publication/3091143 85_Fast_obstacle_distance_estimation_using_laser _line_imaging_technique_for_smart_wheelchair [4] https://www.researchgate.net/publication/3037668 17_Improved_Optical_Flow_Estimation_In_Wron g_Way_Vehicle_Detection [5] https://www.researchgate.net/publication/3048080 65_Evaluation_of_Roadside_WrongWay_Warning_Systems_with_Different_Types_of _Sensors. [6] https://www.researchgate.net/publication/3224127 50_Automatic_road_accident_detection_technique s_A_brief_survey [7] https://ankitshah009.github.io/accident_forecasting _traffic_camera [8] https://www.geographyandyou.com/disaster/disast er-events/track-troubles-road-accidents-in-india/ [9] https://www.researchgate.net/publication/2714499 69_The_investigation_of_car_movement_trajector ies_after_contact_in_the_case_of_collisions_with_ stationary_safety_barriers [10] https://timesofindia.indiatimes.com/india/90deaths-n-roads-due-to-rash-drivingncrb/articeshow/61898677.cms