Aedes Science Gateway: A Machine-learning Based Mosquito Egg Counting System

Each year, approximately 400 million people are infected with an arboviral disease from the bite of an Aedes mosquito. Scientists have been using mosquito traps that allow for identifying and counting Aedes mosquito eggs for decision-making on insecticide spraying and abatement. Aedes project is an effort to automatically count the eggs in the trap paper images utilizing computer vision and machine learning techniques. In this study, we developed a new method that can count the eggs automatically and efficiently run on a CPU or a mobile phone. Specifically, we extract patches from the egg images and trained a convolutional neural network (CNN) based classifier to identify the patches as single eggs, egg clusters, or debris. Finally, we count the eggs by the number of these patches. We achieve 92% accuracy in our testing dataset, 1% lower than ResNet50. However, our model takes < 1 second to process one image on a CPU, much faster than ResNet50. Moreover, we build a science gateway for Aedes with a web portal hosting egg collections and a backend service based on our method for counting the eggs.