Deep Learning Hybrid Model for Predicting Grassy Shoot Disease in Sugarcane

Scientists from the Sugarcane Research Institute claimed that in Uttar Pradesh, India, grassy shoot disease can be found among the different varieties of sugarcane crops; the incidences vary from 6 to 28 percent. So, the estimation is made here that the incidence of the disease is highly prevalent and could probably hamper the growing of sugarcane. Traditional prediction and identification techniques are overwhelming and costly for collecting complete data, while geographic information can be less detailed and the datasets may be small. This paper has considered sugarcane to develop an optimal prediction and classification system of grubby shoot disease using a recently used deep learning algorithm known as CNN-SVM in image classification and recognition with high accuracy. These CNNs were trained on a large dataset of images of sugarcane, which the model would process in extremely high precision to delineate slight differences and features that might have escaped the notice of other conventional methods. This study accomplished an amazing 96% accuracy rate in forecasting grassy shoot disease, a fact that underlined the effectiveness of CNNs in enhancing the forecast of disease and classification.