Mushroom identification and classification are critical areas of research due to the significant health risks posed by poisonous varieties. Poisonous mushrooms present a considerable threat to public safety as they can be easily mistaken for edible varieties, potentially leading to severe poisoning or even death. The lack of accessible and reliable resources for accurately distinguishing between edible and poisonous mushroom species could result in a growing number of fatalities and health complications within the population. Furthermore, the process of mushroom classification itself is inherently time-consuming, demanding a substantial investment of resources and a comprehensive understanding of mycology. To address these issues, this study aims to develop a mushroom detection prototype specifically for identifying poisonous mushrooms using a mobile application. The application leverages a Convolutional Neural Network (CNN) algorithm to accurately classify mushrooms based on user-submitted images. CNN is one of the deep learning algorithms that is well known for its good performance in image recognition and classification. There are 3 main phases of the research methodology, which cover the data collection and preprocessing, model design and implementation, and performance evaluation. In this study, the developed model achieved a good accuracy of 89%, indicating acceptable performance in distinguishing between edible and poisonous mushrooms. This good accuracy underscores the model's reliability and effectiveness in real-world applications, making it a valuable tool for ensuring public safety.
Amirul et al. (Tue,) studied this question.