Diseases in plants are a crucial factor in determining the plant’s quality. The decrease in crop yield is primarily caused by the presence of infected crops, resulting in a reduction in the production rate. Image recognition plays a crucial role in the domain of intelligent agriculture, particularly in the agricultural field. The investigation of early-stage plant disease identification remains unexplored. This study utilizes four deep learning models, namely DenseNet121, ResNet152V2, InceptionV2, and MobileNetV2, to accurately detect the disease called ‘False Smut’ in Rice crops. To achieve improved quality and productivity, it is crucial to promptly detect False Smut. This paper utilizes a unique dataset which is obtained by gathering images of diseased crops from rice fields. The DenseNet121 model achieved an accuracy rate of 98.75% through the implementation of the Transfer Learning training method.