Development of a deep learning-based classification method for agricultural crops using digital images
Abstract
This study concentrates on the difficulty associated with the automated identification and classification of agricultural goods by analyzing digital imagery. For the scope of this inquiry, a deep learning framework, namely a neural network composed of numerous artificial layers, was engineered. This framework enables the extraction of important visual attributes through sequential convolutional operations and then establishes the precise type of crop using fully connected portions of the network. The testing stage utilized a dataset containing images representing thirty-one unique types of cultivated flora. The results confirmed that the developed model demonstrates sufficient accuracy in classifying agricultural crops. The research findings have practical significance for implementing digital technologies in agriculture, monitoring crops, and managing productivity processes.