A Deep Learning-based Approach for Edible, Inedible and Poisonous Mushroom Classification

Abstract

Mushroom is one of the fungi types' food that has the most powerful nutrients on the plant. Nevertheless, the identification of edible, inedible and poisonous mushrooms among its existing species is a must due to its high demand for peoples' everyday meal and major advantage on medical science. For this purpose, deep learning approaches like InceptionV3, VGG16 and Resnet50 has been applied to identify the mushrooms based on their category on 8190 mushrooms images where the ratio of training and testing data was 8:2. Contrast limited adaptive histogram equalization (CLAHE) method has been used along with InceptionV3 to obtain the highest test accuracy. A comparison has been evaluated between contrast-enhanced and without contrast-enhanced method. Finally, InceptionV3 has achieved 88.40% accuracy which is the highest among the rest implemented algorithms.