Deficiencies Identification and Detection in Grape Plant Leaves Using LSTM and CNN Model

To address mineral shortages in soils, modern fertilization methods identify nutritional deficits in plant foliage. To formulate a fertilization plan that delivers essential micronutrients without excessive production of others, it is crucial to ascertain the requisite quantities of vitamins and minerals. Nutritional inadequacies in agriculture are signaled by the leaves of plants. Physical examination of these leaves from the plant's morphological components provides sufficient information to determine nutritional deficiencies. The main aim of this study was to ascertain nutritional deficits in grapevine leaves. This research use Convolutional Neural Networks (CNNs) for model training. By inputting the model's output into an LSTM, we can categorize grape plant leaf images as healthy or exhibiting deficiencies in nitrogen, iron, phosphorus, manganese, potassium, zinc, calcium, boron, magnesium, or sulfur. Accuracy evaluates the model's efficacy. The data analysis indicated that the proposed model achieved an accuracy rate of 98.6 percent. The results indicate that convolution neural networks (CNNs) may be effective in detecting nutritional deficiencies in grape plant leaves.