Development of a model for recognition of wheat rust diseases using hybrid segmentation and deep learning techniques

Abstract

Wheat is one of the most common staple crops and every year, a million tons have been newlineexported from India worldwide. Among all states of India, more than 50% of wheat grew in newlinedifferent regions of Punjab. According to the Indian council of agricultural research, every year newlinemore than 10% growth rate of the wheat crop has decreased due to wheat rust diseases (WRD). newlineWheat rust diseases are caused by the Puccinia fungus pathogen and are one of the most newlinecommon threats in wheat production. The rust diseases encompass stem, stripe, and leaf rust newlinevariants, each posing a significant threat to both yield quality and grain quantity. The rust newlinediseases are identified by farmers who have no experience that leads to incorrect identification newlineresulting in significant losses. Thus, the monitoring of the wheat rust diseases is important that newlinedecreases the grain yield and production quality losses. A CVBA is important for rust disease newlineidentification. that can detect rust leaf lesions timely and precisely. Recently deep learning and newlineimage segmentation techniques have shown significant success by accurately classifying and newlinelocalizing the prediction of various types of rust diseases. In this thesis, three different hybrid newlinemodels of image segmentation and deep learning are proposed for the classification of rust newlinedisease leaf lesions. These models are simulated on various wheat plant images to accurately newlineidentify rust disease classes. newlineThe first proposed model is based newline