« Video-Based Algorithms for road accident detection » par Mme Boutheina MAALOUL

Quand ?
Le 16 juillet 2018 À 10:00
Où ?
Université de Valenciennes (UVHC)

Organisé par

Secrétariat des études

Promoteurs

  • Pr Carlos VALDERRAMA, UMONS/Faculté Polytechnique/Sce SEMi – Belgique
  • Pr Smail NIAR, LAMIH Université de Valenciennes – France

Résumé

Automatic video surveillance systems have been developed to detect and analyze abnormal behavior or situation of risk in many fields reducing human monitoring of activities captured by cameras (security surveillance, abnormal behavior detection, etc.). One of applications of video surveillance is traffic monitoring such as traffic flow detection, vehicle tracking and vehicle motion analysis. In particular, motion analysis in roads aims to detect abnormal traffic behavior and sudden events, especially in case of Emergency and Disaster Management.

Road accidents can cause serious injuries affecting mostly the head and the brain, leading to lifelong disabilities and even death; each additional rescue minute can mean the difference between life and death as revealed by The golden Hour [Lerner et al., 2001]. Therefore, providing rapid assistance for injuries is mandatory. Moreover, if not addressed promptly, accidents may cause traffic jams, eventually leading to more accidents, and even greater loss of life and properties.

Indeed, traffic cameras provide detailed information on the accident causes, a direct communication infrastructure, and video storage resources to be used as evidence for further analysis.  Many cities in France are equipped with video surveillance cameras installed on different roads and highways. Traffic monitoring is done by human operators to visualize the congestion of a road or to measure the flow of traffic. The video stream of this existing network of cameras is delivered raw and unprocessed to the traffic management center and is not registered. Thus, there are no video storage of accident scenes. In addition, there is no associated technology for emergency management. It is therefore important to design a system capable of effectively organizing emergency response respecting the Golden hour criteria. This response should be based on automatic detection and rapid notification of road accident and optimization of the safety intervention itinerary, without affecting the traffic state.

The objectives of the thesis are firstly the identification of accident scenarios and the collection of data related to road accident. Next, the design and development of video processing algorithms for the automated detection of accidents in a one way road. The solutions developed will use the existing fixed cameras, so as not to require significant investments in infrastructure. The core of the proposed approaches will focus on the use of the dense Optical Flow (OF) algorithm [Farnebäck, 2003] and heuristic computations for features extraction and detection. The purpose of the dense OF is to estimate the velocity of each pixel in the region of interest (ROI) between two given frames. At the output of the dense OF, there is a dense features which is more performant than features extracted at some points. Defining thresholds for accident detection in various environment is very challenging. Therefore, studying the motion at a global scale in the image, allows defining a dynamic thresholds for accident detection using statistic computations. The proposed solution is sufficient and robust to noise and light changes.

Liste des publications

  • International Conferences with program committee and proceedings
    • Maaloul Boutheina, Taleb Ahmed Abdelmalik, Niar Smail, Harb Naim, Valderrama Carlos, « Adaptive Video-Based Algorithm for Accident detection on Highways » in « International Symposium on Industrial Embedded Systems », SIES Toulouse, France (2017)
    • Maaloul Boutheina, Taleb Ahmed Abed Al-Malek, Niar Smail, Valderrama Carlos, Derraz Foued, Harb Naim, « Vision-based Traffic Accident Detection Techniques Opportunities and Challenges » in The 2015 International Conference on Advanced Communication Systems and Signal Processing (2015)
    • G.Fritz, B.Maaloul, V.Beroulle, O.Aktouf, D.Hély. “Read rate profile monitoring for defect detection in RFID Systems”. IEEE International Conference on RFID-Technologies and Applications (RFID-TA 2011), Sitges, Barcelona, Spain, on September 15-16, 2011).
  • National Workshops
    • B.Maaloul, E.Fabiani, L.Lagadec. “Modélisation outillée d’architectures reconfigurables supportant les blocs IP matériels”. Colloque nationale Gretsi 2013, Brest, France. Septembre 3-6, 2013.
    • B.Maaloul, E.Fabiani, L.Lagadec. “ Evolution d’une chaine d’outils pour le prototypage d’architectures reconfigurables tolérantes aux fautes ”. Colloque nationale GDR SOC-SIP 2013, Lyon, France. Juin 10-12, 2013.