Scribd
Upload
14 views

Plant Detection

Machine learning and its implementation

Uploaded by

Chonchol Ahmmed
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
14 views

Plant Detection

Machine learning and its implementation

Uploaded by

Chonchol Ahmmed
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 19

Project presentation on

Crop disease detection using


ML and IoT
Department of Electronics & Communication Engineering
SDM Institute of Technology, Ujire
1
Team Members

Project Guide
Mr. Amith K Jain
2
Contents
Problem statement
Objectives
Literature review
Proposed Methodology
Proposed timeline
Approximation expenditure
Expected outcome
Conclusion

3
Problem statement
• Loss of crops affects livelihood

• Presently, there isn’t any scientific proven way to accurately detect a

crop disease
• Detection is purely based on past experience
• Reaching everyone physically seems very complex

• Lack of knowledge and expertise in the field of farming

• Problem for knowledge transfer to younger generation

• Crop disease detection using ML and IoT has been taken up as problem
statement
4
Rationale behind Problem statement

• Problem is prevalent, happening and as of now there isn’t a viable


solution

• Farmer must wait for days to know the kind of disease and possible solution

5
Objectives

To develop a viable product which accurately detects


crop disease

To design a model which assists in early detection of


crop diseases such as
• Potato Early Blight
• Tomato yellow leaf
• Pepper bell Bacterial spot

To integrate ML and IoT in arriving at a viable


solution for the existing problem

6
Literature Review(Background work)
• Early detection of crop using Histogram & Edge detection
1 • Use of ML technique for proper decision regarding disease in
crop

2 • Designed a system for early & accurate disease detection


• Reduce manual efforts for detecting diseased crops

(1)Anuradha Badage, International Research Journal of Engineering and Technology (IRJET)


https://www.irjet.net/archives/V5/i9/IRJET-V5I9158.pdf

(2)Suyash S. Patil and Sandeep A. Thorat , Second International Conference on Cognitive Computing
and Information Processing (CCIP) https://ieeexplore.ieee.org/abstract/document/7802887

7
Proposed methodology
Proposed methodology
Use Raspberry Pi and
Pi-cam to get the
Result will be displayed on the image of the plant
screen by which the end user Circuit design to set the
will be able to know whether parameters
the crop is healthy or not

Result by the ML
algorithm will be sent
ML algorithm will be
back to Raspberry pi from
defined to identify
the cloud
whether the plant is
infected by disease or not
Image captured by
Raspberry Pi will be Python programming
analyzed by trained ML language will be used
algorithm being
deployed at cloud

8
First phase deals with training of data sets - training
data sets for both healthy and diseased crops, using a
convolutional neural network

ML works in 2
Phases

Second phase deals with self evaluation comprising


of identifying the disease

9
Proposed block diagram

10
• Recommended input voltage is 5V, and the recommended input
Pow current is 2A
er • Proposed power supply design is 5.1 V and 2.5 A
supp
ly
• Camera module of raspberry pi clicks an image by pressing the
button
Pi • Image will be sent to the cloud storage where the Machine Learning
cam model is being deployed
era

• It acts as a gateway between the user and Machine learning model


Ras • It takes the image as an input from the user and sends it to the cloud,
pber where the proposed ML model performs crop disease detection and
ry- gives result
pi
11
• This includes training both healthy as well as diseased data sets
Machi • It monitors the crop identifying the disease using a
ne convolutional neural network algorithm from the image sent by
Learni the user
ng
Model

• It assists in displaying the result of the crop image captured and


Micro to display the result / disease
HDMI
Displa
y

12
Data sets collected so far

13
Sample Images from Datasets
Pepper Bell Healthy Pepper bell Bacterial spot

Potato Healthy Potato Early Blight

14
Discussions carried out

17th sept 2020 • Discussion about the necessary


requirements for project

9th Oct 2020 • Discussion on synopsis and the data sets

• Discussed on the algorithm/s to be used


13th Oct 2020 in ML

15
Expenditure
Sl.No. Item Cost(INR)

1 Raspberry pi and Camera Module 3000

2 Display 300

3 Miscellaneous 3000

4 Total approximated cost 6300 16


Expected outcome

Early and accurate detection of healthy


and diseased crop

Integration of ML and IoT to arrive at a


better solution than what is existing

17
Conclusion
Proposed project model aims at early and accurate detection of
the crop disease; same to be displayed on the Micro HDMI
display

Information transfer to next generation as part of integrating


technology and farming

Time required to detect a disease and arrive at a possible


solution will be less; Geographical disadvantage will be
eliminated

18
Thank you

19

You might also like