Isabelo D. Paat Jr.
ABSTRACT
The cultivation of mushrooms plays a vital role in the global food industry and agriculture. However, their susceptibility to various diseases poses a significant threat to yield and quality. To address this challenge, this study presents an innovative approach for automating the detection and classification of mushroom diseases using Convolutional Neural Networks (CNNs). A diverse dataset comprising images of healthy and diseased mushrooms from various species was utilized. A pre-trained CNN architecture was fine-tuned to leverage its feature extraction capabilities, enhancing the model’s ability to detect subtle patterns and characteristics associated with different diseases. The proposed system incorporates transfer learning, data augmentation, and extensive hyperparameter tuning to optimize performance. Experimental results demonstrate that the CNN-based approach accurately identifies and classifies mushroom diseases, achieving state-of-the-art accuracy. The model shows strong potential as a reliable tool for early disease detection in mushroom cultivation. Furthermore, real-time deployment possibilities were explored, offering practical implications for farmers and stakeholders in the mushroom industry. This research contributes to the field of agricultural technology and highlights the potential of deep learning techniques in addressing critical challenges in food production and crop management. As mushroom cultivation continues to expand globally, the automation of disease detection through CNNs holds great promise for improving yield and ensuring the sustainability of this essential food source.
Keywords: Agricultural technology, automated disease detection, convolutional neural networks (CNN), deep learning techniques, data augmentation, hyperparameter tuning, transfer learning
https://doi.org/10.57180/ngmd7490