Automatic identification and behavioral analysis of phlebotomine sand flies using trajectory features

The present paper reports an automated approach for the characterization and analysis of the behavioral of sand flies; the method used is based on Gaussian mixture model and Kalman filter for the detection and tracking of sand flies, and then the extraction of an optimized set of features from the t...

Full description

Saved in:
Bibliographic Details
Published in:The Visual computer 2019-05, Vol.35 (5), p.721-738
Main Authors: Machraoui, Ahmed Nejmedine, Diouani, Mohamed Fethi, Mouelhi, Aymen, Jaouadi, Kaouther, Ghrab, Jamila, Abdelmelek, Hafedh, Sayadi, Mounir
Format: Article
Language:English
Subjects:
Algorithms
Artificial Intelligence
Artificial neural networks
Automation
Behavior
Cameras
Classification
Computer Graphics
Computer Science
Datasets
Embedded systems
Flies
Image Processing and Computer Vision
Insects
Kalman filters
Original Article
Probabilistic models
Support vector machines
Tropical diseases
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The present paper reports an automated approach for the characterization and analysis of the behavioral of sand flies; the method used is based on Gaussian mixture model and Kalman filter for the detection and tracking of sand flies, and then the extraction of an optimized set of features from the trajectory of flight is performed for the classification process. So, we propose here two optimized sets of features; the first one is used to identify sand flies among other insects, and the second is employed for the characterization of the behavioral change in the sand flies in the presence of a repulsive odor. These features are tested on three different classifiers; artificial neural network, support vector machine and K-nearest neighbor (KNN), and the results show an important improvement in the classification accuracy and confirm the effectiveness of our approach; the accuracy rate of the proposed method reached 88.6% for the identification of sand flies and 93.4% for the detection of their behavior change. Instead of the excessive use of pesticides over wide areas, the presented investigation is a key pillar of the development of an ecological way for a statistical information gathering about sand flies in order to fight against disease carried by those insects especially leishmaniosis and pappataci fever.
ISSN:0178-2789
1432-2315
DOI:10.1007/s00371-018-1506-x