ABSTRACT
Agriculture underpins all economies. Crop output affects any nation's local market. Plant diseases reduce productivity after equipping fields with effective resources. Early plant disease type identification is needed for destruction avoidance and detection. Recent study has identified plant diseases that affect several crops worldwide.Many efforts are made to determine these illnesses‘ causes. Some diseases are viral, whereas others are fungal. It becomes a serious issue when the root cause cannot be located before it affects a large percentage of the production process. This study compares computer vision and image methods for plant disease identification and classification. Segmentation pre-processes pictures for plant disease identification. Numerous feature extraction and categorization methods enable it. Model training uses BiGRU, BiGRU-LSTM, and BiGRU-AM. Recommended approach BiGRU-AM works well.
