Intelligent Surveillance System For Detection Of Potential Mosquito Breeding Sites

Abstract

Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) have revolutionised contemporary progress by creating a significant impact across society. Technological advancements in ML and DL have elevated the allure and applicability of AI across diverse areas. The array of applications for ML and DL algorithms is extensive, with their utilisation in solving intricate problems in various domains such as cybersecurity, healthcare, agriculture, and banking. newlineVector control plays a pivotal role in public health management in India, considering the high prevalence of vector-borne diseases in the country. According to the World Health Organization, in 2019, India reported around 3.38 million malaria cases, accounting for 88% of patients in Southeast Asia. Furthermore, dengue, with 111,880 reported cases and 56 deaths in 2019, and Japanese Encephalitis, with 5,425 points and 694 deaths in the same year, remain significant public health challenges. These diseases, transmitted primarily through mosquitoes, present a health risk and a considerable economic burden due to healthcare costs and productivity loss. Therefore, effective vector control is vital in India to reduce the disease burden, protect public health, and improve economic conditions. This importance underlines the need for robust surveillance, public awareness campaigns, and innovative approaches, such as leveraging technology in vector control strategies. Clean water, road surface monitoring, and mosquito breeding site detection are essential aspects of human life and infrastructure management. This work presents the creation of two image datasets for stagnant water and wet surface detection, detection of surface monitoring across different seasons. The datasets comprise 1,976 labelled images for water and wet surfaces and 8,484 photos, and 10 videos for road surfaces. A novel approach using anchor boxes in the improved YoloV3 model is proposed to increase the accuracy of detecting various types of stagnant water