Loading content...
Plant diseases are key constraints to agriculture
and directly impact on crop yield and food security. Early and
reliable diagnosis of the disease is crucial to increase agriculture
production and minimize economic loss. This thesis proposes a
Hybrid Intelligent Model for Precise Plant Disease Detection
and Agricultural Productivity Improvement using deep learning
combined with optimization. The model is a fusion of deep
learning and optimization processes; in which the convolutional
neural networks perform an automated feature extraction,
while the classifier and feature selection are learnt by a selected
algorithm. Image pre-processing techniques are utilized for
enhancing the quality of data, after which a hybrid learning
based model is used to effectively identify diseases in several
crops. Experimental results reveal that, In terms of accuracy,
precision, recall, F1-score and RMSE the proposed model
outperforms traditional image processing methods, machine
learning techniques and the individually trained deep.
Agriculture is a vital component of the global economy
and a primary source of food security. Plant diseases, on the
other hand, continue to be a major danger to agricultural
productivity, resulting in large yield losses[1] and economic
consequences on an annual basis. Traditional illness detection
approaches, such as manual inspection and expert diagnosis,
are frequently time-consuming, labor intensive, and subject to
human error. In recent years, the application of intelligent
computer techniques in agriculture has emerged as a possible
answer to these problems. Hybrid intelligent models, which
incorporate different machine learning and deep learning
methodologies, have shown very promising results in terms
[2]of disease detection accuracy and timely intervention. Plant
disease detection includes identifying symptoms, classifying
disease types, and estimating severity. Image-based analysis,
aided by computer vision, has become the most popular
method for automatic detection since it allows for non
invasive and speedy monitoring. Convolutional Neural
Networks (CNNs)[3] have been widely employed for feature
extraction and classification due to its capacity to detect subtle
spatial patterns in leaf pictures. However, single-model