Field Effectiveness of Drones to Identify Potential Aedes aegypti Breeding Sites in Household Environments from Tapachula, a Dengue-Endemic City in Southern Mexico

Abstract

Aedes aegypti control programs require more sensitive tools in order to survey domestic and peridomestic larval habitats for dengue and other arbovirus prevention areas. As a consequence of the COVID-19 pandemic, field technicians have faced a new occupational hazard during their work activities in dengue surveillance and control. Safer strategies to monitor larval populations, in addition to minimum householder contact, are undoubtedly urgently needed. Drones can be part of the solution in urban and rural areas that are dengue-endemic. Throughout this study, the proportion of larvae breeding sites found in the roofs and backyards of houses were assessed using drone images. Concurrently, the traditional ground field technician’s surveillance was utilized to sample the same house groups. The results were analyzed in order to compare the effectiveness of both field surveillance approaches. Aerial images of 216 houses from El Vergel village in Tapachula, Chiapas, Mexico, at a height of 30 m, were obtained using a drone. Each household was sampled indoors and outdoors by vector control personnel targeting all the containers that potentially served as Aedes aegypti breeding sites. The main results were that the drone could find 1 container per 2.8 found by ground surveillance; however, containers that were inaccessible by technicians in roofs and backyards, such as plastic buckets and tubs, disposable plastic containers and flowerpots were more often detected by drones than traditional ground surveillance. This new technological approach would undoubtedly improve the surveillance of Aedes aegypti in household environments, and better vector control activities would therefore be achieved in dengue-endemic countries.

A Drone System with an Object Identification Algorithm for Tracking Dengue Disease

“…Regarding the technical characteristics that a drone must have to adapt an epidemiological surveillance system, the drone selected in the methodology is evaluated, the DJI MINI 2 drone, which offers greater stability despite being very light (250gr), it has an acceptable wind resistance (29 -38 km/h), however in [4] it mentions a study that used the DJI Phantom 4 Pro drone for epidemiological surveillance of malaria in the Peruvian Amazon, identifying breeding sites through multispectral images captured by the drone, said drone has an autonomy greater than 30 minutes, in the same way in [7] they used the same drone to capture images in an endemic city of Tapachula in Mexico. Likewise, in [6] the use of the DJI Mavic Pro Platinum drone is mentioned to obtain images of a test area, which has a flight time of 30 minutes, considered very high, it also has GPS, a 12.35 MP camera, among other features.…”

“…Regarding the development of the database, in [7] a database acquired by the drone of a total of 2579 images of various house roofs was used to identify Aedes Aegypti breeding sites, divided into 10 categories. On the other hand, in [15] they use Microsoft's COCO data set, which contains a base of 330,000 images, divided into 91 categories that can be freely used by any user.…”

“…With respect to the simulation, images containing the objects to be detected were used, for example buckets, tubs, bottles, pots, vases, tires, etc., which were extracted from the internet. Similarly in [7], for the detection of dengue it is recommended to identify these objects.…”

“…For this reason, an image of an area of the Najran University campus in Saudi Arabia was captured using a DJI Mavic Pro Platinum drone. Similarly in [7] , they mention that the use of unmanned aerial vehicles (UAV) provides the option of collecting detailed spatial information in real-time at a relatively low cost and thus avoids limitations associated with satellite data (such as cloud pollution, low resolution, bad camera angle and shooting time for taking pictures). Therefore, in their inspection of breeding habitats through drone surveillance, they used a lowcost Phantom 4 Pro DJI model drone, which integrates a 20-megapixel camera with a focal length of 35 mm and a theoretical resolution of 1 cm , which has a battery for a flight of 12 to 15 minutes, which allowed them to take photos of 15 to 20 houses per day.…”

Unmanned aerial vehicles for surveillance and control of vectors of malaria and other vector-borne diseases

“…Water that is muddy or covered by vegetation may be missed by visible-spectrum sensors [37], however hyperspectral sensors can detect water bodies by their thermal signature [36]. The detection of mosquito breeding sites with UAVs has also been investigated in urban environments, with the mobility of drones used to observe rooftops that may otherwise be inaccessible to technicians [38]. However, the acceptability of UAVs to local communities is important to assess prior to their deployment in the field, with cross-sectional studies of the general population undertaken in Malaysia, Mexico, and Turkey [28] to assess public views.…”