Leveraging Swarm Optimization with Deep Learning Driven Nutrients Deficiency Diagnosis in Rice Crop Management

Rice (Oryza sativa L.) is the fourth main food crop globally and is crucial for food security worldwide.Fertilization plays a significant part in rice yield and excellence and is a major portion of rice field management.Nutrient deficiency in the soil is considered as the main factor.Fundamentally, macronutrients are nutrients that contain more attention than micronutrients and are vital for tissue growth and plant cells.The most essential nutrients for plants are nitrogen (N), phosphorus (P), and potassium (K).A computer vision-based automated nutrition position of rice recognition model has occurred in agriculture.Deep convolutional neural network (DCNN) learning depends upon an artificial neural network (ANN) that can absorb and create intelligent forecasts using techniques.This objective focuses on the design of the Rhinopithecus Swarm Optimization Algorithm for Diagnosing Nutrient Deficiency Rice Crop (RSOA-NDRC) technique for the identification of nutrient deficiency in rice crops using a parameter-tuned Deep Learning (DL) model.Several stages of the RSOA-NDRC technique achieve this.In preprocessing, RGB images are converted into HSV images to remove the background.Binary images are produced based on hue and saturation to separate the diseased from the non-diseased areas.In addition, Feature extraction is performed using the ShuffleNet method.The deep fuzzy neural network (DFNN) approach was applied to identify nutrient deficiencies.Finally, the rhinopithecus swarm optimization (RSO) model-based hyperparameter selection procedure is used to enhance the recognition outcomes of the DFNN approach.A comprehensive set of experimentations was performed to establish the improved performance of the RSOA-NDRC technique.The performance validation of the RSOA-NDRC technique portrayed a superior accuracy value of 98.85% over existing models.