Shivi

https://doi.org/10.64823/ijter.2608002

Plant disease is one of the primary issues that directly reduce the quality of agricultural produce. Identification and classification of plant diseases are among the primary tasks to improve the overall quality of crop production for economic development. Numerous approaches for disease detection have been demonstrated by a group of researchers using a digital image dataset of plant leaves. However, due to their intricate algorithms and incapacity to provide an effective method for clearly demarcating boundaries among the provided data classes for the purpose of making the ultimate decisions, such systems fall short of the anticipated perfection. Deep Neural networks, which are multi-layer architecture models whose layers learn to represent the data at several abstract levels and have a less complex algorithm compared to conventional statistical methods, can be suggested as a solution for this problem. In the present context of the problem, a self-designed Multi-Branch Fusion Network model for the multiclass categorization of plant diseases has been proposed in this work. Images from several plant diseases were taken from a publicly accessible dataset and used to simulate the proposed model. It has been discovered that the suggested method works well in the specified situation.