Image-Based Macroscopic Classification of Aspergillus Fungi Species Using Convolutional Neural Networks

Abstract

This paper presents a technique for macroscopic classification of Aspergillus fungi species. The Aspergillus genus have several species that can be used in agricultural and medical applications. An automated process of macroscopic identification and classification of such species is described here. The scope of the study includes a 9-type Aspergillus fungi species. The learning mechanism used is a simple convolutional neural network. Using a total of 4545 macroscopic images, the model achieved a 90.06% accuracy in training, and 96.43% accuracy in validation.