Soil health agricultural by drones imagery, pixilation and machine learning
This article proposes an agricultural computer system which is the subject of a patent application. It describes a method of studying the soil from drone image based on machine learning techniques in order to give results explosible directly by farmers. This article defines the different technologies used namely the drone image classification by cosine similarity and the supervised learning by a feedforward neural network. Key Words: Artificial Intelligence - Agriculture - Computer Vision - Normalized Difference Vegetation Index - Drone
READ FULL TEXT