Real-Time Mosquito Species Identification using Deep Learning Techniques
Pratik Mulchandani1, Muhammad Umair Siddiqui2, Pratik Kanani3
1Pratik Mulchandani, Department of Computer Engineering, Dwarkadas J. Sanghvi College of Engineering, University of Mumbai, India.
2Muhammad Umair Siddiqui, Department of Computer Engineering, Dwarkadas J. Sanghvi College of Engineering, University of Mumbai, India.
3Pratik Kanani, Department of Computer Engineering, Dwarkadas J. Sanghvi College of Engineering, University of Mumbai, India.
Manuscript received on November 22, 2019. | Revised Manuscript received on December 15, 2019. | Manuscript published on December 30, 2019. | PP: 2000-2003 | Volume-9 Issue-2, December, 2019. | Retrieval Number: B2929129219/2019©BEIESP | DOI: 10.35940/ijeat.B2929.129219
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: According to the World Health Organization, diseases such as malaria and dengue account for almost one million deaths every year. Carrier mosquitoes for a particular disease remain exclusive to it. A majority of carrier mosquitoes spread the disease throughout a region by reproducing in it. With advancements in Machine Learning and Computer Vision technologies, the species of mosquitoes in a particular region can be easily and swiftly detected using recordings of their wing movements. The wingbeats of a particular mosquito species are unique, making this a reliable method to identify them. Once these solutions are deployed on mosquito traps, a particular region can be alerted if, for example, an Aedes Aegypti mosquito is found. This mosquito species is widely known to carry the Zika virus. The identification of such carrier species can also help in detecting the spread of mosquito-borne diseases in the surveyed region. In this paper, we go through various techniques that show promising results in the identification of mosquito species. The trained models can be deployed on constrained devices to make a cost-effective and efficient mosquito species identification system.
Keywords: Carrier mosquitoes, constrained devices, machine learning, mosquito detection, deep learning