IFATA‐Deep net: Improved invasive feedback artificial tree algorithm with deep quantum neural network for root disease classification

Abstract

SummaryIdentification of crop plant disease in the early phase is one of the most vital tasks in agriculture. The infection by disease resulted in huge loss to economy. Thus, quick and precise treatment of disease prevented the product loss and enhances the product quality. In this work, a method for categorizing the root disease is developed. The goal is to create a model utilizing a collection of photos of roots to categorize root sickness. Initially, noise is removed during the pre‐processing using a Gaussian filter. The segments are generated using the Pyramid Scene Parsing Network (PSPNet). Here, PSPNet training is carried out using the improved invasive feedback artificial tree method (IFATA), which is developed by combining the improved invasive weed optimization (IIWO) and Feedback Artificial Tree (FAT). Data augmentation is done to make an image suitable for further processing. Root disease is categorized using a deep quantum neural network. With the suggested IFATA, Deep Quantum Neural Network (DQNN) is trained. The analysis of technique is performed with two databases, namely Rice root Gellan dataset and Alfalfa root crowns. The proposed IFATA‐based DQNN outperformed with higher accuracy, sensitivity, and specificity scores of 93.5%, 94.5%, and 90.5%, respectively.