Automated Disease Detection in Oyster Mushroom Cultivation Using Deep Learning Technique

Mushroom cultivation is highly dependent on controlled environmental conditions, and fungal diseases pose a significant threat to the maintenance of crop quality and yield. Traditional methods for identifying diseases, such as manual inspections, are time consuming and error prone and often do not provide timely results. To overcome these challenges, this study proposes a real-time image-based classification system using the YOLOv5 object detection model. The system categorizes mushrooms into two groups: healthy and diseased. By processing image data in real-time, YOLOv5 facilitates accurate and prompt detection of fungal diseases, enabling timely intervention to minimize crop loss and improve yields. The proposed model leverages the efficient feature extraction and classification capabilities of YOLOv5, achieving exceptional performance in experimental trials. The system demonstrated a recall of 93 % and a precision of 94 % to detect diseased mushrooms, resulting in an F1 score of 84%. For healthy mushrooms, it achieved a precision of 88%, a recall of 85 %, and an F1-score of 86 %. The general precision of the model in classifying healthy and diseased mushrooms was 88 %. This innovative approach integrates advanced image processing and machine learning techniques, providing a scalable and effective solution for sustainable mushroom farming while reducing labor dependency and improving productivity.