APPLICATION OF FASTER R-CNN MACHINE LEARNING MODEL TO DETECTE EDIBLE AND NON-EDIBLE MUSHROOMS

Abstract

Mushrooms have long been recognized as a nutritious food source, rich in antioxidants and fiber. They are also a good source of B vitamins, selenium, zinc, and copper - essential elements for cellular energy production and a strong immune system. As a result, mushrooms have become a popular culinary ingredient worldwide. However, accidental consumption of poisonous mushrooms can lead to severe consequences, including nausea, vomiting, neurological impairment, disorders, acute anemia, and even death if not treated promptly. Distinguishing between poisonous and edible mushrooms is not always easy, as their appearances can be remarkably similar. This paper aims to address this issue by utilizing the Faster Region-based Convolutional Neural Network (Faster R-CNN) machine learning model to classify edible and poisonous mushrooms. The Faster R-CNN model is trained on a diverse dataset of mushroom images, focusing on features such as shape, color, and texture. Following the training process, the model achieved an impressive accuracy of 99.10% in mushroom classification. This result demonstrates the potential of Faster R-CNN in aiding users to identify safe mushrooms, contributing to the reduction of mushroom poisoning and related fatalities.