Automated detection and tracking of phlebotominaes

Ahmed Nejmedine Machraoui (1) , Mohamed-Fethi Diouani (2) , Jamila Ghrab (3) , Mounir Sayadi (1)

1 SIME-ENSIT - Laboratoire Signal, Image et Maitrise de l’Energie [Tunis]
2 LR11IPT03 - Laboratoire d’Epidémiologie et de Microbiologie Vétérinaire
3 Laboratoire de Parasitologie, Institut Pasteur de Tunis

This paper presents a method for detecting and tracking phlebotominaes that are vectors for many important diseases. A method based on the Gaussian mixture model is used for Foreground/Background classification. Then, the mathematical morphology operations are used to refine the classification results and eliminate areas that do not represent phlebotominaes. Thereafter, the Kalman filter is used for the prediction and stimation of their positions, thus the Hungarian algorithm is used for the assignment of the detected positions to tracks. The results can help us study the behavior of these insects, and to improve traps and solutions that limit their spread in case of infection. The proposed method is tested on a real phlebotominaes video sequences recorded at Pasteur institute of Tunisia, and the experimental results show the efficiency of the proposed algorithm to accurately detect the position and track phlebotominaes.

Keywords: Gaussian mixture model; Kalman filter; Hungarian algorithm; Multi-objects Tracking; Motion detection

Domains: Life Sciences [q-bio]

Ahmed Nejmedine Machraoui, Mohamed-Fethi Diouani, Jamila Ghrab, Mounir Sayadi. Automated detection and tracking of phlebotominaes. ATSIP 2014 : 1st International Conference on Advanced Technologies for Signal & Image Processing, 2016, ⟨10.1109/ATSIP.2016.7523069⟩. ⟨pasteur-01451893⟩