Aedes aegypti mosquito movements analysis and sex classification using computer vision and deep learning

Mosquitoes are vectors of diseases, carrying viruses, parasites, and bacteria that infect millions of people around the world. Understanding their flight patterns and behaviours is crucial for disease modelling, ecological research, and developing effective control methods. Traditional manual methods for analysing mosquito flight records are time-consuming and limited, whilst automated methods could provide a viable alternative. In this study, the recognition, monitoring, and classification of mosquito movements was done using artificial intelligence (AI), particularly, computer vision and deep learning. Two experiments were carried out: the first experiment assessed the system's capacity to accurately detect and classify the direction of mosquito movements using various classifiers, with models such as Gated Recurrent Unit (GRU) and Convolutional Neural Network model with Long Short-Term Memory model (CNN-LSTM). Results show a high accuracy rate of 96.67%. The second experiment showed the system's ability to identify between male and female Aedes aegypti mosquitoes using a CNN model based on movement heatmaps. Results show accuracy levels between 89.84% and 99.73%.