| Abstract | Maize is one of the largest food crops after rice and wheat in all over the world. It has become
a vital part in the agriculture to automatically monitor and diagnosis the plant disease in new
era. In this paper, a unique maize plant disease identification system has been modelled using
the optimization of recurrent neural network through bacterial foraging technique. Further it
can be employed to moderate the search space of the classifier and provides better initial
information for recognition network. Initially pre-processing and region growing based
segmentation has been approved to identify and determine the neighbour pixel of seed points
toward combining it in particular region. It is to increase the efficiency of the network by
searching and grouping of seed points having common attributes to generate the optimal
features for classification. Obtained segmented region has been processed using Recurrent
Neural Network to capture sequential patterns based on longitudinal data to produce recurrent
structures. Finally different classes of the plant disease have been classified and identified with
new classic techniques. Experimental analysis using plant village dataset on the proposed
model explains the performance measure in rudiments of dice coefficient, similarity and
specificity towards disease identification.
|
|---|
