AN ADVANCED AI-POWERED PLATFORM FOR FARMERS: MASTERING THE ART OF PLANT DISEASE DETECTION, MONITORING, AND PREDICTIONS

Abstract

The growing global population needs food and resources from agriculture. One of the biggest challenges for farmers is plant diseases. These diseases reduce crop yields and productivity. Identification, tracking, and forecasting are necessary to control these diseases and prevent losses. Farmers' inability to identify crop plant diseases is the main issue. Farmers without medical expertise may not be able to distinguish diseases due to their similar symptoms. Misdiagnosis can lead to ineffective or inappropriate treatments that waste resources and worsen the problem. After a disease outbreak is confirmed, tracking and forecasting should prevent its rapid spread. Farmers traditionally identified plant diseases by observation and experience. This method is subjective, time-consuming, and error-prone. Knowledge, experience, and disease symptom recognition are essential for farmers. Some farmers consult agricultural experts or extension workers, but this is rarely scalable due to cost and availability. Innovative AI-driven and cloud-based platforms for farmers are needed to overcome traditional methods' limitations. This platform would improve plant disease tracking, identification, and forecasting with AI, cloud computing, and data analysis. This platform helps farmers increase crop yields, reduce losses, and improve global food security with cutting-edge technology, knowledge, and real-time insights

Specification

Description:According to the facts, training and testing of DL-CNN involves in allowing every source
image via a succession of convolution layers by a kernel or filter, rectified linear unit
(ReLU), max pooling, fully connected layer and utilize SoftMax layer with classification
layer to categorize the objects with probabilistic values ranging from . Figure 1 discloses
the architecture of DL-CNN that is utilized in proposed methodology for CBIR system for
enhanced feature representation of word image over conventional retrieval systems.
Convolution layer as depicted in Figure 1 is the primary layer to extract the features from a
source image and maintains the relationship between pixels by learning the features of image
by employing tiny blocks of source data. It's a mathematical function which considers two
inputs like source image where and denotes the spatial coordinates i.e., number
of rows and columns. is denoted as dimension of an image (here , since the source
image is RGB) and a filter or kernel with similar size of input image and can be denoted as
.
The output obtained from convolution process of input image and filter has a size of
, which is referred as feature map. An example of
convolution procedure is demonstrated in Figure 3.2. Let us assume an input image
with a size of and the filter having the size of . The feature map of input image is
obtained by multiplying the input image values with the filter values as given in Figure 1.
An electronic device intended to support plant wellbeing through the vibe of the plant, and
visual side effects can be helpful for novices in the horticulture framework as well as
prepared specialists as a confirmation apparatus in sickness medication. Progresses in PC
5
vision offer a chance to expand and foster exact plant security checking and to widen the
interest for PC vision applications in the field of accuracy cultivating. A well-known
computerized picture process method, for example, variety investigates and limits [10] were
utilized to recognize and order plant infections. There are different methodologies for word
related criminal investigator infections, and most styles are fake brain organizations (ANNs).
They are joined with marginally various methods of pre-handling the picture as far as
removing higher elements. The cerebrum is included many profoundly interconnected
neurons cooperating to take care of issues. A human-produced nerve cell can be a piece of a
cycle with different information sources and one result. With each info, the nerve cell
frequently has loads that are connected with a general predisposition. Proposed framework
has layered module as follows Input Layer, Convolution Layer, Activation Function Layer,
Pool Layer, and Fully Connected Layer. Input Layer: This layer contains the picture's crude
contribution to width 32, level 32 and profundity.
Convolution Layer: This layer works out the volume of the result by ascertaining the speck
item between all channels and fixing the picture. Assume we utilize a sum of 12 channels to
get yield volume of aspect 32 x 32 x 12 for this sheet. Actuation Function Layer: This layer
applies component wise initiation capability to convolution layer execution. Nearly an
enactment capabilities incorporate RELU: max (0, x), Sigmoid: 1/(1+e^-x), Tanh, Leaky
RELU, and so on. Pool layer: interaction of lessen picture into chosen aspects (for example
5x5 or 2x2 to apply the aspect lattice. Completely associated layer associated with all layers.
, C , C , Claims:We claim the platform's AI algorithms can identify plant diseases with up to 95%
accuracy, significantly reducing the risk of misdiagnosis.
2. We claim Farmers can track the health of their crops in real-time, enabling them to
respond to potential threats immediately, minimizing crop loss.
3. We claim the platform can predict potential disease outbreaks up to two weeks in
advance, allowing farmers to take proactive measures to protect their crops.
4. We claim By optimizing the use of pesticides and fertilizers based on real-time data, the
platform can help farmers reduce input costs by up to 20%.
5. We claim the platform is designed to be intuitive and easy to use, requiring no advanced
technical knowledge, making it accessible to farmers of all skill levels

Documents