ABSTRACT

Areca nut, sometimes called betel nut, is a tropical crop. Areca nuts are produced and
consumed in India, which is the world's second-largest producer and consumer. It suffers
from a range of ailments during its life cycle. Farmers use their senses of sight to detect
disease. Multiple image processing techniques for categorization of Areca nuts with various
properties such as colour, texture are examined in this project. Computer detection systems
have been widely employed in the real world for retrieval tests because they can provide
rapid, efficient, accurate, and clear testing. Until now, areca nut separation has been done by
hand. Areca nut separation employing a complex colour sorting mechanism is made up of a
variety of exterior properties of the nut, such as colour, texture, form, and size. Various
approaches will be used to extract information from the captured image. The areca nut's
colour can also be used to classify it. To efficiently plan the areca nut, all of these traits are
essential. To extract composition characteristics, the co-occurrence matrix of wavelet
coefficients at the second level of details, as well as contour let coefficients, is used.
Separation is accomplished by mixing three texture elements namely strength, contrast, and
homogeneity. This approach has been shown to work effectively in the contour let domain,
allowing for a reduction in the vector dimension feature. When given training data, it
categorises the data into healthy and unhealthy areca nuts based on colour and quality.

i
CHAPTER

INTRODUCTION

i
 Areca nut is one of India's most important commercial plants. Currently, areca nut sorting is done by hand, which
is a time-consuming and inefficient process. We'll need to segregate automatically for this. Machine Learning is a
hidden AI platform that may be utilised to automate this procedure. The website is created to classify areca nuts
according to their quality. When it comes to grading, the division of the areca nut is important. Sorting correctly
aids in determining the appropriate yield. CNN is an In-depth Learning method that can capture an image, give
value (readable weight and bias) to the various elements in the image, and differentiate between them. When
compared to other classification methods, CNN requires extremely little prior processing Despite the fact that the
filters are constructed by hand in traditional methods, CNN can interpret these filters / symbols with the right
training.

1.1 Problem Statement

Classification of crops is one of the important processes in precision agriculture. Classification of crops based on
their verity, enhances the quality. In this project, we presented a study of three main supervised classifiers, Yolo is a
efficient classifier for classifying the raw arecanut using color histogram and color moments as features.

1.2 Objectives

 This project mainly aims at developing a computer vision based efficient grading system for boiled
Arecanuts. Grading of Arecanut is a tedious process in Arecanut marketing.
 The Main objective is to classify arecaunuts into categories like bette , hassa etc…

1.3. Scope:
We present a system where image analysis studies aiming at automated detection of disease that may be present in
areca nut by using images of areca nut and bunches. In this work, we propose a convolution image processing
model that has been trained with images of areca nut.

i
CHAPTER 2
LITERATURE SURVEY

In order to get required knowledge about various concepts related to the presentapplication, existing literature were
studied. Some of the important conclusions weremade through thoseare listedbelow.

Siddesha S. et al., [1] this interacts with the division of the green areca nut based on histogram color and color times
with KNN separators. The experiment was developed utilising the KNN algorithm and an 800-photographic
database of four classes with two-color characteristics and four-grade scales. A separate image is utilised in the
second stage to extract elements. Color histogram and colour timing approaches were used to extract this colour
information. The areca nut is divided into four classes in the third stage. KNN, ANN, and SVM are three well-
known category designers that we employ to divide.
Dhanuja K.C. et al., [2] advocated using imaging processing technology to identify areca nut disease, and the author
employed grading based on areca nut stitching. The machine's viewing programme was created to recognise and
identify areca nodes as distinct, dignified, and poor levels. The Deep Learning algorithm and imaging techniques
are used to classify the data.
Anil kumar MG et al., [3] in this article, we focus on early detection of diseases in Areca nut using convolution
neural networks. CNN is an in-depth Learning algorithm used to diagnose diseases of areca nut, which takes inputs
as a picture, provides readable weights and variations objects in the picture, and then learn from the results to
distinguish one from the other.
Dinesh R et al., [4] Areca nut used image processing and computer vision to partition and manipulate the data. The
author distinguishes the areca nut category based on colour, size, and texture. The major goal of this article is to
give a comprehensive review of the areca nut, Computer Vision, and the technological needs and applications based
on the notion of areca nut categorization and grading.
Bharadwaj N K et al., [5] is proposed to determine the distance of the areca nut through pictures of areca nut. The
proposed future extraction method uses the local binary pattern method. Supports a vector classified machine used
to determine areca nut range. The accuracy, memory, and F -measure generated in the confusion matrix are all used
for testing and grading system performance.
Ajit Danti et al., [6] Raw areca nut separation procedures and methods have been proposed. The author has devised
a novel method for classifying areca nuts into two groups based on their colour. Segmentation, Masking, and
Classification are all steps in the classification process. The classification is based on two different colours: red and
green areca nut phases in the region. Test performance success rates ranged from 97 to 98 percent depending on the
category.
Mallikarjuna S. B. et al., [7] proposed a CNN-based algorithm for classifying various diseases pictures of areca nut.
Separate the various diseases such as rot, split and rot. Results from a four-phase data set shows that the proposed
strategy excels in terms of classification, memory, accuracy, and F steps.
i
Hubert Cecotti. et al., [8] have found grape clusters using the neural convolution network. A separate classification
techniques used to separate two grape varieties of white and red grapes at the same time Use the CNN algorithm to
perform 99% accuracy.
Ashish Nage et al.,[9] this paper focuses on a method based on image processing to find plant diseases. Here is the
paper the author made the helpful Android app growers in diagnosing plant diseases by uploading a leaf image to a
program using CNN leaf-cutting algorithm.
Prasanna Mohanty et al., [10] has proposed how to diagnose plant diseases by training a CNN. The CNN model is
trained to identify 14 healthy and diseased plants. The model achieved 99.35% accuracy in test set data. When using
a model in pictures purchased from reliable online sources, the model achieves 31.4% accuracy, while this better
than a simple random selection model, a very different set of training data can help to increase accuracy. And some
variation of model or neural network training may produce high accuracy, thus opening up a way to make the
diagnosis of plant diseases more easily accessible for everyone.
Ashfaqur Rahman et al., [11] presented an image processing and artificial intelligence process for locating mature
areca nut. The entire process is divided into two steps: the first step separates the areca nut images from the
background image, and the second step classifies the areca nut. Image based on the mature group and achieved an
overall accuracy of 96 percent.
Hubert Cecotti et al. [12], A convolution neural network was used to recognise grape bunches. The CNN algorithm
was used to segment the two types of grapes, white and red grape bunches, using several segmentation techniques.
It performed with 99 percent accuracy.
Scarlett Liu et al., [13] suggested an image-based grape bunch detecting algorithm. The grape bunches are counted
from photos using an image processing and support-vector machine technique, with an overall success rate of 88
percent for red grapes.
Lucas Mohimon et al., [14] have been compared deep-learning and machine-learning algorithms for grape bunches
cluster segmentation. There were 200 photos of white grapes in typical light in the collection, and the accuracy was
86 percent. When compared to traditional segmentation techniques, the results demonstrated that deep-learning
algorithms are more robust in detecting white colour grape bunches.
Ashfaqur Rahman et al., [15] To locate the mature grape bunches, a series of image processing and intelligence
processes were applied. The complete procedure is divided into two steps: the grape bunches photos are removed
from the background image in the first phase, and the grape bunches are classified based on the mature group in the
second step, with an overall accuracy of 96 percent.
Alharbi et al., [16] Different diseases on apples have been identified and classified. On testing images, the
classification accuracy was 90%.
Ajit Danti et al., [17] proposed raw arecanut segmentation and classification algorithms. The author has devised a
revolutionary method for classifying arecanuts into two colour groups. Segmentation , Masking and Classification.
The classification is based on the red and green components of the segmented section of the arecanut. In the
experimental efficiency, the categorization success rate was around 97 to 98 percent.

i
Ajith Danti et al., [18] suggested a system for grading arecanuts that is both efficient and effective. For successful
arecanut segmentation, three sigma control limits on colour features were created.
Ajit Danti et al., [19] described techniques for rating arecanuts based on textural qualities. Classification is based on
mean around features, GLCM features, and combination (Mean around-GLCM) features. A decision tree classifier
is used to divide arecanuts into six groups. GLCM and Mean Around-GLCM features were employed by the
Decision Trees classifier. The cross validation approach is used to do the testing. The experimental results show that
GLCM features are 97.65% successful, Mean Around features are 98.28 percent successful, and Mean Around-
GLCM features are 99.05 percent successful.
Sameer Patil et al., [20] Pre-processing techniques should be used to classify the grade of arecanuts, according to
the author. A Raspberry Pi system with a 5 Megapixel camera module was used to take the photos. Image filtering,
contrast enhancement, and an image segmentation algorithm are used to pre-process the captured image. Canny
edge detection and the K-means segmentation algorithm were used to locate the nut's boundary while cropping the
arecanut images. To offer the best results for the arecanut segregation method in this study, the author categorised
eight alternative picture pre-processing methods.
AjitDanti et al., [21] proposed an arecanut grading system. For improved results, the author combined two
algorithms, KNN and SUV, to classify the nuts.
Dhanesha R. et al., [22] The HSV colour model was used to determine and identify the automated maturation stage
of arecanut bunches. 200 photographs of arecanut bunches at varying stages of ripeness were used in the
experiments.
Dhanesha R. et al., [23] determined the maturity degree of arecanut bunches. The YCgCr colour model
automatically splits an arecanut bunch from a given image, allowing the maturity level of the arecanut to be
determined. The author produced a database of 1000 photos using a smartphone camera. The testing result
demonstrates that this method can accurately link input images with an accuracy of 80%.
Dhanesha et al., [24] showed dynamic contouring segmentation in which arecanut bunches in pictures were
segmented. The author photographed 20 different arecanut bunches at various stages of ripeness. VOE and DSC
performance measures were used to examine the data's accuracy.
Siddesha et al., [25] Oil palm crop bunch photos were segmented using trained and unsupervised methods. The
author used Hill climbing, Grow cut, Random Walker, MSRM algorithms for supervised techniques and K-Means,
Fuzzy-CMeans algorithms for unsupervised techniques on a dataset of 100 bunch pictures. The impact would be
better if the same approaches were used on a bunch of arecanuts.
Yashima Ahuja et al., [26] presentedthe concept of Multiclass classification and many classification approaches
were discussed. Support Vector Machines has a good record in Binary classification. The study of fitness of SVM in
multiclass classification is done in this paper.
T. Liu et al.,[27] paper is about the problem solving of betel nuts by manual ways, the automatic classification
method on the basis of computer vision is discussed in this paper.

i
S.R. Kodituwakku et al., [28] determines which color or texture is better to represent the similarity of colors images.
Color moment and color histogram descriptors are not effective since image pixels aren’t into considerations. CCV
and co-occurrence matrix features shows more effective images other than color and texture feature.
A.Danti et al., [29] paper is about the classification techniques of arecanut based on text feature is done. GLCM and
mean around GLCM features, based on this classification is done. Six classes classification is done based on
Decision tree.
André Silva Aguiar et al., [30] proposed employing several image-processing and deep-learning models to
recognise the grape bunch at various phases of development. The collection contains 1929 photos of the bunch at
various phases

i
CHAPTER 3
SYSTEM REQUIREMENTS
HARDWARE REQUIREMENTS

 System : intel i3/i5 2.4 GHz.

 Hard Disk : 500 GB
 Ram : 4/8 GB

SOFTWARE REQUIREMENTS

 Operating system : Windows XP/ Windows 7/8/10.

 Packages Used : Open CV Python
 Coding Language : Python
 Toolbox : Image processing toolbox.
 IDE :Python

Functional Requirements:

 System should detect the Image with arecaunut.

 System should detect and classify the arecaunuts.
 System should automatically display the result.

i
Non Functional Requirements:

Usability

 Easy Interface for capture of image and classify the arecaunuts

Reliability

 System should be reliable enough to classify the images

Performance

 Should not take excessive time in classifying

Supportability

 Contain easy to understand code with provisions for future enhancement.

i
CHAPTER 4

DESIGN AND ARCHITECTURE

Systems design is the process of defining the architecture, components, modules, interfaces, and data for a system to
satisfy specified requirements. Systems design could see it as the application of systems theory to product
development. System design is one of the most important phases of software development process. The purpose of
the design is to plan the solution of a problem specified by the requirement documentation. In other words, the first
step in the solution to the problem is the design of the project. The design of the system is perhaps the most critical
factor affecting the quality of the software. The objective of the design phase is to produce overall design of the
software. It aims to figure out the modules that should be in the system to fulfil all the system requirements in an
efficient manner.

System Architecture

i
SEQUENCE DIAGRAM

The figure 5.2 shows a sequence diagram, simply depicts interaction between objects in a sequential order i.e the
order in which these interactions take place. We can also use the terms event diagrams or event scenariors to refer to
a sequence diagram.Sequence diagrams describe how and it what order the objects in a system function.

Figure 5.3: Sequence diagram

i
ACTIVITY DIAGRAM

The figure 5.3 shows an Activity diagram, which describes the dynamic aspects of the system. It is a flowchart
that describes the flow from one activity to another. The flow can be sequential, branched or concurrent.
Activity diagrams are not only used for visualizing the dynamic nature of a system, but they are also used to
construct the executable system by using forward and reverse engineering techniques.

START

EXTRACT
IMAGE

PREPROCESSING

FEATURE SELECTION
AND FEATURE

CLASSIFICATION

CLASSIFY

STOP

Figure 5.4: Activity diagram

i
CLASS DIAGRAM

The figure 5.4 shows a class diagram, it is the main building block of object-oriented modeling. It is used for
general conceptual modeling of the structure of the application, and for detailed modeling translating the models
into programming code. Class diagrams can also be used for data modeling. The classes in a class diagram represent
both the main elements, interactions in the application, and the classes to be programmed.

CAMERA

UPLOAD THE PREPROCESS CLASSIFY THE

IMAGE MODULE MODULE

CLASSIFY AND ANALYSE THE

CATEGORIZE MODULE

Figure 5.5: Class diagram

i
Use Case Diagram

A Use Case Diagram is a lot of situations that reflect a client-frame relationship. A use case chart shows the
entertainer-to-use relationship. Usage cases and on-screen characters are the two main elements of an usage case
diagram. An on-screen character refers to an user or other person connected with the demonstrated process. A use
case chart in figure is an out-of - the-box perspective that speaks to some activity each module will perform to
complete an errand.

Fig : USE CASE DIAGRAM

Data Flow Diagram

i
The figure 5.5 shows a Data flow diagram is a way of representing a flow of data through a process or a system.The
DFD also provides information about the outputs and inputs of each entity and the process itself. A data flow
diagram has no control flow, there are no decision rules and no loops.

Level: 0 describes the overall process of the project. We Input the image of the arecaunt .

System Classifies the arecaunut Based on Yolo

Level 2:
i
  Level 2 describes how Arecaunut is Classified
 System Classifies Arecaunut based on Data sets Trained and Extracted Features

Figure 5.6: Data Flow diagram

CHAPTER 6
i
 METHODOLOGY

In this chapter, the algorithm for detection and classification of arecaunut is discussed in detail. YOLO object
detection algorithm is used for detection and classification. Here yolo workflow is discussed in step by step.

Yolo – Object Detection Algorithm:

Deep Learning consists of a very enormous number of neural networks that use the multiple cores of a process of a
computer and video processing cards to manage the neuranetwork’s neuron which is categorized as a single node.
Deep learning is used in numerous applications because of its popularity especially in the field of medicine and
agriculture. Here YOLO deep learning technique is used to identify persons wearing and not wearing face masks.
Joseph Redmon et al. introduced You look only once also known as YOLO in 2015. YOLO is a convolutional
neural network (CNN) for doing object detection in real-time.

i
The algorithm applies a single neural net5 work to the full image, and then divides the image into regions and
predicts bounding boxes and probabilities for each region. These bounding boxes are weighted by the predicted
probabilities. Some improvements were done over years and YOLOv2 and YOLOv3 , v5 versions were introduced
respectively in 2016 , 2018.Our model uses YOLOv5 and it provides good results regarding object classification
and detection. In the previous version of Yolov3 Darknet-19 is used. Yolov3 uses darknet-53. Darknet is a
framework used for training neural networks written in C language

Benefits
YOLO is a popular object detection algorithm because it achieves high accuracy while it is also able to run in real-
time. The algorithm “only looks once” at the image means that it requires only one forward propagation pass
through the neural network to make predictions. After non-max suppression it then gives the recognized objects
along with the bounding boxes. In YOLO, a single CNN simultaneously predicts multiple bounding boxes and class
probabilities for those boxes. YOLO directly optimizes detection performance since it trains on full images. YOLO
has a number of benefits over other object detection methods they are
• YOLO is extremely fast
• YOLO scans the entire image during training and also during testing. So, it implicitly encodes contextual
information about classes as well as their appearance.
• YOLO learns generalizable representations of objects so that when it is trained on natural images and tested , the
algorithm performs excellently when compared to other top detection methods.

Configuration Module
i
First, we created a method to connect with user’s local network of cameras through internet or use the internet
accessible IP cameras. For our solution we have made use IP cameras but we can make use of other types cameras
as well with slight changes to the code. Each camera has its own instance and it will leverage all the features
discussed in this paper and process for each instance will be the same as shown in Fig (1). To connect cameras to
the application, users have to provide their camera IPs/or anything equivalent to it like RTSP URL, etc. These IPs
will be saved in database for reuse.

Classification Using Yolo

We use the IPs stored in the database as input for object detection module and we use OpenCV for any image
processing related tasks. For object detection we are using pretrained YOLOv5 model trained on COCO dataset
because of its high accuracy. As showed by [8] YOLOv5 is better than previous proposed algorithms hence it is our
preferred choice of object detection model. YOLO recognizes objects more precisely and faster than other
recognition systems. It can predict up to 80 classes. It can be easily trained and deployed in a production system.
YOLO is based on a single Convolutional Neural Network (CNN). The CNN divides an image into regions and then
it predicts the boundary boxes and probabilities for each region. It simultaneously predicts multiple bounding boxes
and probabilities for those classes. YOLO sees the entire image during training and test time so it implicitly encodes
contextual information about classes as well as their appearance. The real-time recognition system will recognize
multiple objects from an image. The backbone of YOLOv5 is CSPDarknet-53. Darknet is an open-source neural
network framework written in C and CUDA and it is used to train new YOLO models and run the existing ones.
After the training is done the results are stored in a .weights file. We have converted this darknet model which is
in .weights format to a TensorFlow model in .pb format. We use this saved model for object detection. YOLOv5
sends us the coordinates of detections along with confidence score which we then use as input for OpenCV to draw
bounding boxes on the detections and count these detections. We count detections on every frame and the process
repeated for each frame. We also able to selective counting with help of detection boundary which discussed below.

Data Annotation

i
To train YOLO we need to annotate images for object detection models. Our dataset should be well annotated.
There are different types of annotations available. Here a bounding boxes method is used. It creates a rectangle area
over images that are present in our dataset. Since Annotation needs more time we are using a tool called LabelIMG
to annotate our data. 3.3.3 YOLOv5 Configuration The YOLOv5 configuration involved the creation of two files
and a custom Yolov5 cfg file.

YOLOv5 configuration

first creates a ”obj.names” file which contains the name of the classes which the model wanted to detect. Then a
obj.data file which contains a number of classes in here is 2, train data directory, validation data, ”obj.names” and
weights path which is saved on the backup folder. Lastly, a cfg file contains 2 classes. figure 3.2 shows the
configuration steps involved. Next is training of our YOLOv3 in which an input image is passed into the YOLOv5
model. This will go through the image and find the coordinates that are present. It divides the image into a grid and
from that grid it analyzes the target objects features. Here 80 percent data is used for training , and remaining 20
percent is used for validation. Now weights of YOLOv5 trained on the dataset are created under a file. Using these
trained weights now we can Detect and count the crowd.

Algorithm (Step by Step Execution)

i
Step 1: Start the program.

Step 2: Input image is loaded .

Step 3: YOLOv5 trained weights are loaded from the disk.

Step 4: arecaunut Detected and marked by means of object detection algorithm.

Step 5: After detection resultant image is displayed.

Step 6: Classify the Type.

Yolo V 5 Training:

There is an existing implementation using YOLOV5 in github however the model is incomplete and also doesn’t
support additional architectures due to which is not published. We used this as reference and built a YOLOv5 social
distancing model which is capable of working on multiple architectures used in the YOLOv5 model as well on our
custom architecture (modified backbone CSP). By doing this, we have successfully implemented the first working
social distancing model based on YOLOv5 and other supporting architectures based on YOLOv5.

Implementation of Social Distancing on YOLOv5 Models Our key objective is to implement a working YOLOv5
based social distancing model using YOLOv5s, YOLOv5s6, and YOLOv5s6 modified bottleneck CSP architecture.
There is no existing social distancing working model based on the above YOLOv5 architectures. From the
screenshots below the model’s sample output can be seen which clearly indicates the risk category based on
centroid distance calculations. Parameters for High, Medium, and Low risk: High Risk: Distance between people
less than 200 units. Medium Risk: Distance between people between 200 - 250 units. 11 Low Risk: Distance
between people more than 250 units.

Modules:
i
 Capture the image
 Identify the arecaunut from the image
 Recognise the arecaunut.
 Classify

PseudoCode

1. Capture the Image.

2. Read the video by dividing it into a number offrames.
3. Else, detect persons in each frame and get the bounding boxes around them with the help
ofYOLOv5.
a. If it reaches to EOF,stop.
.
4. While detecting arecaunut, detect their shape , color size.
5. With the help of bounding boxes on the arecaunut .

6. Create a results board on top of the image and display the results.
7. Create an output stream and then show theresults.
8. Do this for every frame till it reaches to end offile.

TESTING
Testing is the process of evaluating a system or its component(s) with the intent to find whether it satisfies the
specified requirements or not. Testing is executing a system in order to identify any gaps, errors, or missing

i
requirements in contrary to the actual requirements.

 TestingPrinciple

Before applying methods to design effective test cases, a software engineer must understand the basic principle
that guides software testing. All the tests should be traceable to customer requirements.

 TestingMethods
There are different methods that can be used for software testing. They are,
1. Black-BoxTesting

The technique of testing without having any knowledge of the interior workings of the application is called
black-box testing. The tester is oblivious to the system architecture and does not have access to the
sourcecode.Typically,whileperformingablack-boxtest,atesterwillinteractwiththesystem'suserinterface by providing
inputs and examining outputs without knowing how and where the inputs are workedupon.

2. White-BoxTesting
White-box testing is the detailed investigation of internal logic and structure of the code. White-box
testingisalsocalledglasstestingoropen-boxtesting.Inordertoperformwhite-boxtestingonanapplication,
atesterneedstoknowtheinternalworkingsofthecode.Thetesterneedstohavealookinsidethesourcecode and find out
which unit/chunk of the code is behavinginappropriately.

Levels ofTesting

There are different levels during the process of testing. Levels of testing include different methodologies that
can be used while conducting software testing. The main levels of software testingare:
i
  FunctionalTesting:

This is a type of black-box testing that is based on the specifications of the software that is to be tested.
The application is tested by providing input and then the results are examined that need to conform to
the functionality it was intended for. Functional testing of software is conducted on a complete,
integrated system to evaluate the system's compliance with its specified requirements. There are five
steps that are involved while testing an application forfunctionality.

 The determination of the functionality that the intended application is meant toperform.

 The creation of test data based on the specifications of theapplication.

 The output based on the test data and the specifications of theapplication.

 The writing of test scenarios and the execution of testcases.

 The comparison of actual and expected results based on the executed testcases.

 Non-functionalTesting

This section is based upon testing an application from its non-functional attributes. Non-functional
testing involves testing software from the requirements which are non-functional in nature but important
such as performance, security, user interface, etc. Testing can be done in different levels of SDLC. Few
of them are

Unit Testing

Unit testing is a software development process in which the smallest testable parts of an application, called
units, are individually and independently scrutinized for proper operation. Unit testing is often automated but it
can also be done manually. The goal of unit testing is to isolate each part of the program and show that
i
individual parts are correct in terms of requirements and functionality. Test cases and results are shown in the
Tables.

i
 Arecaunut Classification Using Yolo

Unit Testing Benefits

 Unit testing increases confidence in changing/ maintainingcode.

 Codes are morereusable.
 Development isfaster.
 The cost of fixing a defect detected during unit testing is lesser in comparison to that of
defects detected at higherlevels.
 Debugging iseasy.
 Codes are morereliable.

Unit testing:

Sl # Test Case : - UTC-1

Name of Test: - Image or video capture

Items being tested: - Input Image

Sample Input: - Camera Stream

Should Capture input image

Expected output: -

Actual output: - Image Captured Successful

Remarks: - Pass.

Page 26
Arecaunut Classification Using Yolo

Sl # Test Case : - UTC-2

Name of Test: - Object Detection

Items being tested: - Labelling

Sample Input: - Image or video

Objects like Helmet , Mask , gun

Expected output: -

Actual output: - Objects Detected

Remarks: - Test Passed

Sl # Test Case : - UTC-3

Name of Test: - Arecaunut Detection

Items being tested: - Different Mask images

Sample Input: - Image or video

Arecaunut should be detected

Expected output: -

Actual output: - Same as Expected

Remarks: - Test Passed

Page 27
Arecaunut Classification Using Yolo

Integration Testing:

Integration testing is a level of software testing where individual units are combined and tested as a
group. The purpose of this level of testing is to expose faults in the interaction between integrated
units. Test drivers and test stubs are used to assist in Integration Testing. Integration testing is
defined as the testing of combined parts of an application to determine if they function correctly. It
occurs after unit testing and before validation testing. Integration testing can be done in two ways:
Bottom-up integration testing and Top-down integrationtesting.

1. Bottom-upIntegration

This testing begins with unit testing, followed by tests of progressively higher-level
combinations of units called modules or builds.

2. Top-downIntegration
In this testing, the highest-level modules are tested first and progressively, lower-level
modules are testedthereafter.

In a comprehensive software development environment, bottom-up testing is usually done

first, followed by top-down testing. The process concludes with multiple tests of the complete
application, preferably in scenarios designed to mimic actual situations. Table 8.3.2 shows the test
cases for integration testing and their results.

Page 28
Arecaunut Classification Using Yolo

Sl # Test Case : - ITC-1

Name of Test: - Image Capture and automated arecaunut Detection and Classification

Item being tested: - Arecaunut Detection

Sample Input: - Video Stream or Image

Expected output: - Arecaunut should be detected in Video or image

Actual output: - Classified into different types

Remarks: - Pass.

System testing:

Page 29
Arecaunut Classification Using Yolo

System testing of software or hardware is testing conducted on a complete, integrated system to

evaluate the system's compliance with its specified requirements. System testing falls within the
scope of black-box testing, and as such, should require no knowledge of the inner design of the
code or logic. System testing is important because of the following reasons:

 System testing is the first step in the Software Development Life Cycle, where the
application is tested as awhole.

 The application is tested thoroughly to verify that it meets the functional and
technicalspecifications.

 The application is tested in an environment that is very close to the production environment
where the application will bedeployed.

 System testing enables us to test, verify, and validate both the business requirements as
well as the applicationarchitecture.

System Testing is shown in below tables

Sl # Test Case : - STC-1

Name of Test: - System testing

Item being tested: - OS Compatibility

Sample Input: - Video Stream

Expected output: - Should work in windows , linux

Actual output: - Worked in all os platforms

Remarks: - Pass

CHAPTER 7

Page 30
Arecaunut Classification Using Yolo

RESULTS

ADD PROJECT SCREENSHOTS

CHAPTER 8
Page 31
Arecaunut Classification Using Yolo

CONCLUSION AND FUTURE ENHANCEMENT

CONCLUSION

Arecanut is grown by many farmers in the coastal district, primarily for commercial use. Arecanut is
currently one of the most in-demand and highly regarded crops. Because of the high rate, arecanut has
piqued the interest of nearly all farmers in Uttara Karnataka and the Mysore region. Chemical
fertiliser, chemical liquid spraying, climate changes such as heavy rain, and temperature variations
have all contributed to the sickness of the arecanut Hence Farmers suffer a loss. Every year before the
monsoon, then every 40 days, this practice is followed by all farmers. a long time If the farmers are
aware of the diseases and whether or not the crop is afflicted, Preventing the disease before the
monsoon, followed by a spray, will be effective. Taking care and Controlling bunches using polythene
bags is another option. Our purpose is to educate farmers by providing them with information. The use
of technology to locate and classify illnesses in arecanuts. The focus of our research is on putting in
place a system that uses machine.

Page 32
Arecaunut Classification Using Yolo

REFERENCE

[1].Siddesha, S., Niranjan, S. K., & Aradhya, V. M. (2018). Color Features and KNN in
Classification of Raw Areca nut images. In 2018 Second International Conference on Green
Computing andInternet of Things (ICGCIoT), 1(1), 504-509.
[2]. Dhanuja, K. C., & Mohan Kumar, H. P. (2020). Areca Nut Disease Detection using Image
Processing Technology. International journal of engineering research & technology (IJERT),
9(8),223-226
[3]. Anilkumar M G., Karibasaveshwara TG., Pavan HK., SainathUrankar., Dr. Abhay Deshpande.
(2021). Detection of Diseases in Areca nut Using Convolutional Neural Networks. International
Research Journal of Engineering and Technology (IRJET), 8(5), 4282-4286
[4]. Dinesh, R., & Bharadwaj, N. K. (2017). Possible approaches to areca nut sorting/grading using
computer vision: A brief review. In 2017 International Conference on Computing, Communication
and Automation (ICCCA), 1(1), 1007-1014.
[5]. Bharadwaj, N. K. (2021). Classification and Grading of Areca nut Using Texture Based Block-
Wise Local Binary Patterns. Turkish Journal of Computer and Mathematics Education
(TURCOMAT), 12(11), 575-586.
[6]. AjitDanti., Suresha, (2012). Segmentation and Classification of Raw Areca nuts Based on
ThreeSigma Control Limits, Procedia Technology. Elsevier Ltd, 4(1),215-219.
[7]. Mallikarjuna, S. B., Shivakumara, P., Khare, V., Kumar, V., Basavanna, M., Pal, U., &
Poornima, B. (2021). CNN based method for multi-type diseased areca nut image classification.
Malaysian Journal of Computer Science, 34(3), 255-265.
[8] Hubert Cecottii, H., Rivera, A., Farhadloo, M., &Pedroza, M. A. (2020). Grape detection with
convolutional neural networks. Expert Systems with Applications, 159(1), 1-9
[9]. Mr. Ashish Nage , Prof. V. R. Raut, 2019, Detection and Identification of Plant Leaf Diseases
based on Python, INTERNATIONAL JOURNAL OF ENGINEERING RESEARCH &
TECHNOLOGY (IJERT) Volume 08, Issue 05 (May 2019), [10].Mohanty, S. P., Hughes, D. P.,
&Salathé, M. (2016). Using deep learning for image based plant disease detection. Frontiers in Plant
Science, 7(September), [1419]. [11]. Rahman, A., Hellicar, A. (2014). Identification of mature grape
bunches using image processing and computational intelligence methods, 2014 IEEE Symposium on
Computational Intelligence for Multimedia, Signal and Vision Processing (CIMSIVP), 1(1), 1- 6.
[12]. Cecotti, H., Rivera, A., Farhadloo, M., &Pedroza, M. A. (2020). Grape detection with
convolutional neural networks. Expert Systems with Applications, 159(1), 1-9.

Page 33
Arecaunut Classification Using Yolo

[13].Liu, S., & Whitty, M. (2015). Automatic grape bunch detection in vineyards with an SVM
classifier. Journal of Applied Logic, 13(4), 643-653.
[14].Mohimont, L., Roesler, M., Rondeau, M., Gaveau, N., Alin, F., &Steffenel, L. A. (2021).
Comparison of Machine Learning and Deep Learning Methods for Grape Cluster Segmentation. In
International Conference on Smart and Sustainable Agriculture, 1470(1), 84- 102. Springer, Cham.
[15].Rahman, A., Hellicar, A. (2014). Identification of mature grape bunches using image processing
and computational intelligence methods, 2014 IEEE Symposium on Computational Intelligence for
Multimedia, Signal and Vision Processing (CIMSIVP), 1(1), 1- 6.
[16]. Alharbi, A. G., &Arif, M. (2020). Detection and Classification of Apple Diseases using
Convolutional Neural Networks. In 2020 2nd International Conference on Computer and Information
Sciences (ICCIS), 1(1),1-6
[17].A. DantiArecanut texture-based decision tree categorization, in Proceedings of the CUBE
International Information Conference.
[18].S. PatilAn Areca Nut Pre-processing Algorithm for Quality Classification, in International
Conference on Image Processing and Capsule Networks.
[19].A. Danti et al]Arecanut Grading with an Effective Multiclassifier, in International Conference on
Information Processing. Heidelberg and Berlin: Springer. P. S. Devang, N.
[20].S. Patil, A. Naik, M. Sequeira, G. Naik, and J. Parab (2021). An Areca Nut Pre-processing
Algorithm for Quality Classification, in International Conference on Image Processing and Capsule
Networks.
[21].Danti, A., &Suresha, M. (2012). Effective Multiclassifier for Arecanut Grading. In International
Conference on Information Processing, 292(1), 350-359. Springer, Berlin, Heidelberg.
[22].R. Dhanesha Arecanut Bunch Segmentation Using the HSV Color Model, in the 2018
International Conference on Electrical, Electronics, Communication, Computer, and Optimization
Techniques (ICEECCOT).
[23].R. Dhanesha Arecanut Bunch Segmentation Using the YCgCr Color Model, in 2019 1st
International Conference on Advances in Information Technology (ICAIT).
[24].R. DhaneshaUsing Active Contouring, a Novel Approach for Segmentation of Arecanut
Bunches. In Integrated Intelligent Computing, Communication, and Security. Singapore: Springer.
[25].S. Siddesha Oil palm crop bunch segmentation from tree photos in 2017 International
Conference on Smart Technologies for Smart Nations (SmartTechCon)
[26]. Y. Ahuja, S.K. Yadav. Multiclass Classification and Support Vector Machine, Global Journal of
Computer Science and Technology Interdisciplinary, Volume 12, Issue 11, Version 1.0, 2012, pp.15-
19.

Page 34
Arecaunut Classification Using Yolo

[27]. T. Liu, J. Xie, Y. He, M. Xu and C. Qin, An Automatic Classification Method for Betel Nut
Based on Computer Vision, Proceedings of the IEEE International Conference on Robotics and
Biomimetics, Guilin, China, 2009, pp.19-23
[28]. S.R. Kodituwakku, S.Selvarajah. Comparison of Color Features for Image Retrieval.Indian
Journal of Computer Science and Engineering, Vol. 1 No. 3, 2004, pp.207-211.
[29]. A. Danti and M. Suresha. Texture Based Decision Tree Classification for Arecanut’,
Proceedings of the CUBE International Information Technology Conference, ACM Publications,
2012, pp.113-117.
[30] Aguiar, A. S., Magalhães, S. A., Dos Santos, F. N., Castro, L., Pinho, T., Valente, J., &
Boaventura-Cunha, J. (2021). Grape bunch detection at different growth stages using deep learning
quantized models. Agronomy, 11(9), 1-23.

Page 35