Aedes Mosquito Larvae Recognition with A Mobile App

Abstract

In the era of industrial revolution, mobile application becomes the heart of the intelligent system that integrates Artificial Intelligent (AI) system for autonomous and internet-of-things (IoT). Smartphone acts as an IoT and ubiquitous gadget to perform data analytics for fast detection or prediction. Therefore, the use of the technology is to overcome the problem of increasing number of dengue cases in Malaysia, which the Intelligent Mosquito Larvae Detection Mobile Application (iMOLAP) is proposed in this study. The purpose of iMOLAP is to help the community to responsive about the dengue larvae spotted in their area by using their smartphone and also can be used to classify the species of mosquito larvae. The mobile application uses one of the Convolutional Neural Network (CNN) techniques, which is the Inception V3 model. The new mobile application learns and classify the species of mosquito larvae by referring to a pre-set collection of mosquito larvae species image. The image captured is compared with pre-set image collection to measure the accuracy. As the results, the accuracy shows 92.8% after the image is captured using the mobile application. Finally, iMOLAP successfully analyze and able to classify the aedes species of mosquito larvae from the image taken and detect the affected area of location. The impact of iMOLAP performs fast response in mosquito larvae detection and an awareness tool for the community in combating dengue cases in Malaysia.