Fall armyworm detection in Ghana using machine learning
The economy of many African countries depends on the agriculture sector. However, the agriculture sector is at the mercy of FAW following the arrival of the pest in 2016. The pest is more dangerous to crops than other crop pests because they attack many crops, especially cereal crops, which are the staple food in Ghana. The most widely used technique to control the pest in the country is pheromone-baited traps and manual surveillance. Although these techniques are cheap to implement, they are prone to errors. This paper explores how to use machine learning techniques to detect Fall armyworm (FAW) in Ghana. The paper discusses various data preprocessing methods such as data augmentation, transfer learning, convolutional networks for classification as well as using object detection algorithms to detect the worm.
Undergraduate thesis submitted to the Department of Computer Science, Ashesi University, in partial fulfillment of Bachelor of Science degree in Computer Science, May 2020
fall armyworm (FAW), Convolutional Neural Network (CNN), insect pest recognition, models, models