Abstract:
Potato leaf disease is one kind of venom disease which is the reasons of young
leaves to roll and turn yellow or pink. Early blight, late blight, septoria
blight etc are the familiar types of potato leaf diseases which spectacle
the syndrome on the affected zone and spoil the whole potato gradually.
Appearance of potato leaf diseases depends on weather. These kinds of
disorder syndrome can be cause of qualitiless production and even can slow
down the farmers economy. By detecting this indisposition on primary stage
can bring off farmars from this suffering. Identifying this disorder by experts
is not always affordable on lab specially in rural area. We promote a model
that identify automatically affected leaf conscientiously in a short time . In
this paper, a combined (CNN-LSTM) model along with a traditional machine
learning algorithms employing MR images have proposed to identify disease
on potato leaf. we have used dataset of 12,009 potato leaf images, 7,376
images of them are affected leaf. Heathy & disorder-impacted leaf both were
used to provide a separate layer between abnormal and normal aspects.The
proposed CNN-LSTM model, is succeed in (keras and tensorFlow) and its
ability is surpassing . This Convolutional neural networks(CNN) model can
detect affected potato leaf in short time with the accuracy of 97.78% and
it’s quite reliable than other models.
