One of the main problems that directly lower the quality of agricultural production is plant disease. The primary tasks to enhance the overall quality of plant cultivation for economic development are recognizing and categorizing plant diseases. The same holds true for the rice crop, since approximately 520 million metric tons of rice are used annually. Using a digital image collection of rice leaf datasets, a group of researchers has presented various methods for disease identification. However, such methods don’t achieve the expected perfection because of their complex algorithms and inability to give an effective solution in terms of identifying clear boundaries among the given data classes for making the final decisions. For this problem, deep neural networks (DNN) can be proposed because a DNN is a multi-layer architecture model whose layers learn to represent the data at multiple abstract levels. Further, to learn large databases semantically at a high level and deep feature learning, emerging its applications in object detection and recognition. The present work is the result of a similar motivation, hence proposed a pre-trained model known as Inception_v3 for multiclass’s classification of rice diseases. For the simulation of inception_v3, a publicly available dataset containing images belonging to multiple rice diseases has been used. This dataset contains three diseases belonging to rice leaves known as Bacterial Leaf Blight, Blast, and Brown Spot. The proposed technique is found to be effective in the given context.