Patent Study Report

SYSTEM AND METHOD TO SEGREGATE THE FRUITS

BY MACHINE LEARNING

submitted in

Partial fulfilment of the requirements of the

internship of
Sem/Year: IV/II

Bachelors of Technology in Computer Science & Engineering

Submitted by

Pranita Bhatt(23BT04D190)
Sanskruti Gohil(23BT04D176)
Disha Sarvaiya(23BT04D173)

Under the guidance of

DR. ARCHANA MAGARE

SCHOOL OF
TECHNOLOGY, GSFC
UNIVERSITY
VADODARA – 391750, GUJARAT – INDIA
 TABLE OF CONTENTS

Sr.no Title Page.no

1. Acknowledgement 1
2. Introduction 2
3. Title of Invention 3
4. Field of Invention 4
4. Abstract of Invention 5
5. Patent Requirements 6
6. Different Types of Patents 7
Overview of Utility, Design, and Plant Patents
7. Patent Search Process 9
8. Decoding Patents 10
 Understanding the Patent Lifecycle
 Components of a Patent Application
 Exploring Patent Families

9. Study on patent Claims 12

 Summary
 Analysis of Patent Claims
 Examination of Drawings, Specification, and
Descriptions

10. Future Implications of Patented Product/Process 21

 Forecasting the Future of the Patented Invention
 Potential Impact on Industries and Markets
 ACKNOWLEDGEMENT

Firstly I would like to thank Dr.Sourabh Shah ( Dean ) for meticulously planning
academic curriculum in such a way that students are not only academically sound but
also industry ready by including such industrial training patterns.

I wish to express my sincere thanks and gratitude to my esteemed Faculty

Coordinator Dr.Archana Magare who has contributed so much for successful
completion of REEP Study by their valuable guidance in the field of AI.

Lastly, I would like to thank my team-mates for honest suggestions and giving
me a helping hand in my Industrial training project.

1
 INTRODUCTION

A patent is a special right granted by the government to an individual or organization who

has invented a new industrial product or process. To qualify for a patent, the invention must
meet certain criteria: it must be new, not obvious to others skilled in the field, useful, and
eligible for patent protection according to the country's laws. Essentially, a patent provides
the inventor with exclusive rights to their invention for a limited period of time.

By disclosing their invention to the public through the patent application process, the
inventor is rewarded with legal protection for their creation. This protection allows them to
prevent others from making, using, or selling their invention without permission. Patents
can also be transferred from the original inventor to another party through assignments.

When a patent is granted, it provides territorial rights, meaning that the inventor has
exclusive rights to their invention only in the country where the patent is granted. This
allows inventors to protect their inventions within specific geographical boundaries,
ensuring that they can benefit from their innovation without competition from unauthorized
parties.

2
 TITLE OF INVENTION

System and Method to Segregate the Fruits

By Machine Learning

3
 FIELD OF INVENTION

The present invention pertains to a system and method for handling agricultural products,
specifically fruits and vegetables. More precisely, it concerns a method and system for
segregating agricultural products based on their size and quality, employing a machine
learning mechanism. The background provides context for understanding the invention
without implying that the information constitutes prior art. Fruits and vegetables are valued
for their nutritional content, including protein, vitamins, and fiber. Quality is influenced by
factors like appearance, size, and freshness, leading to grading based on various criteria.
Current grading methods rely on visual inspection and measurement tools, which can be
inaccurate and labor-intensive. Additionally, the system includes provisions for disposing
of low-quality fruits through an openable duct, further streamlining the process.

4
 ABSTRACT OF INVENTION
SYSTEM AND METHOD TO SEGREGATE THE FRUITS BY MACHINE LEARNING
Abstract The present invention relates to segregation of the fruits by machine learning. The
method may include; segregating, the good quality fruits and the average quality fruits from
the received fruits, by utilizing a separator element; segregating, the large size good quality
fruits and the small size good quality fruits from the segregated good quality fruits;
segregating, the large size average quality fruits and the small size average quality fruits
from the segregated average quality fruits; storing, the segregated large size good quality
fruits and the segregated small size good quality fruits, in a first container and a second
container, respectively; storing, the segregated large size average quality fruits and the
segregated small size average quality fruits, in a third container and a fourth container,
respectively.

5
What is the Patent Requirement?

Patent requirements refer to the criteria that an invention must meet in order to be eligible
for patent protection. These requirements typically include:

1. Novelty: The invention must be new and not previously disclosed or known to the
public. It should not have been published, patented, or otherwise made available to
the public before the patent application filing date.
2. Utility: The invention must have a practical use or utility. It should serve a useful
purpose and be capable of being used or applied in some manner.
3. Non-obviousness: The invention must not be obvious to a person skilled in the
relevant field. This means that the invention should not be an obvious modification
or combination of existing technologies or concepts.
4. Enablement: The patent specification must enable a person skilled in the field to
make and use the invention based on the description provided. Sufficient detail and
guidance must be provided to allow others to replicate the invention without undue
experimentation.
5. Patentable subject matter: The invention must fall within statutory subject matter,
which typically includes processes, machines, manufactures, compositions of
matter, or improvements thereof. Not all inventions are eligible for patent
protection; for example, laws of nature, natural phenomena, and abstract ideas are
generally not considered patentable subject matter.

6
Different types of patents
There are several types of patents, each designed to protect different aspects of innovation
and creativity. Here are the main types:

1. Utility Patents: Utility patents are the most common type of patent and are granted
for new and useful processes, machines, articles of manufacture, compositions of
matter, or any new and useful improvement thereof. These patents provide
exclusive rights to the inventor to prevent others from making, using, selling, or
importing the patented invention for a limited period, usually 20 years from the
filing date of the patent application.

2. Design Patents: Design patents protect the ornamental design or appearance of a

functional item. They are granted for new, original, and ornamental designs for
articles of manufacture. Design patents provide protection for the aesthetic aspects
of a product, such as its shape, configuration, surface ornamentation, or a
combination thereof. Design patents typically have a term of 15 years from the date
of grant.

3. Plant Patents: Plant patents protect new varieties of asexually reproduced plants,
such as those produced by grafting or budding. To qualify for a plant patent, the
plant must be novel, distinct, and not obvious to a person skilled in the field. Plant
patents provide exclusive rights to the inventor to reproduce, sell, and use the
patented plant for a period of 20 years from the date of filing the patent application.

4. Software Patents: Software patents protect inventions related to computer software

and algorithms. They cover new and useful processes, methods, or techniques
implemented in software. Software patents can be utility patents if they meet the
requirements for patentability, including novelty, non-obviousness, and utility.

7
 However, the patentability of software inventions varies depending on the
jurisdiction, and some countries have restrictions or limitations on software patents.

5. Biotech Patents: Biotech patents protect inventions related to biotechnology,

including genetically modified organisms (GMOs), gene sequences, pharmaceutical
compositions, and methods for diagnosing or treating diseases.

8
What is Patent Search?
Searching for patents and understanding their lifecycle, application, and family can be
complex, but here's a guide to help you navigate the process:

1. Define Your Search Criteria: Determine the specific technology or invention you're
interested in. This could include keywords, patent classification codes, or inventor names.

2. Choose a Patent Database: There are several patent databases available, including:
● United States Patent and Trademark Office (USPTO): Covers patents granted in the
United States.
● European Patent Office (EPO): Covers patents granted in Europe.
● World Intellectual Property Organization (WIPO): Provides access to international
patents through the Patent Cooperation Treaty (PCT).
● Google Patents, Espacenet, and FreePatentsOnline: Offer free access to a large
database of patents.

3. Perform the Search: Input your search criteria into the chosen database. Use advanced
search features to narrow down results by date, inventor, assignee, classification, etc.
4. Review Results: Examine the patents that match your criteria. Pay attention to patent
titles, abstracts, and claims to determine relevance.

5. Analyze Patents: Once you've found relevant patents, analyze them to understand the
technology, claims, and any prior art they reference.

9
How to decode Patents?

Decoding Patents:
1. Patent Lifecycle:
● Invention: An inventor develops a new technology or product.
● Patent Application: The inventor or assignee submits a patent application to the
relevant patent office.
● Patent Examination: The patent office reviews the application to determine if it
meets the criteria for patentability.
● Patent Grant: If the application meets the requirements, the patent office grants the
patent, providing exclusive rights to the inventor for a specified period.
● Maintenance: The patent holder must pay maintenance fees to keep the patent in
force.
● Expiration: After the expiration of the patent term, typically 20 years from the filing
date, the invention enters the public domain.

2. Patent Application:
● Specification: The patent application includes a detailed description of the
invention, along with drawings (if applicable).
● Claims: Claims define the scope of protection sought by the inventor. They describe
the specific features or elements of the invention.
● Abstract: A concise summary of the invention.
● Filing Date: The date on which the patent application is submitted to the patent
office.

3. Patent Family:
● A patent family consists of all patents filed in various countries to protect the same
invention.

10
● Priority Application: The initial patent application filed in one country, which
serves as the basis for subsequent applications in other countries.
● Divisional Applications: Additional patent applications filed to cover specific
aspects or variations of the original invention.

● Continuation Applications: Applications filed to pursue additional claims or

modifications to the original invention.

11
Study of Patent Claims, Drawings Specification and Description

System and Method to Segregate Fruits by Machine Learning

Summary:
The patent presents a comprehensive system and method for fruit segregation leveraging
machine learning techniques. It addresses the challenges associated with traditional fruit
grading methods by introducing an automated process that enhances accuracy and
efficiency.

Key Points:
1. Segregation Process:
Introduction of a conveyor unit with a separator element to divide received fruits
into two primary categories: good quality and average quality.
Utilization of a bifurcation unit to further classify good quality fruits into large and
small sizes based on predetermined criteria.
Implementation of a similar process for average quality fruits, ensuring precise
segregation.

2. Size Classification:
Incorporation of image sensors at strategic locations to capture images of the
received fruits and analyze their characteristics.

Machine learning algorithms analyze the acquired images to determine the quality
and size of the fruits.

12
 The system identifies large and small size fruits within both good and average
quality categories, enabling precise classification.

3. Storage:
Segregated fruits are stored in designated containers based on their quality and size
classifications. The system ensures efficient distribution and storage of fruits,
optimizing space utilization and facilitating streamlined inventory management.

Benefit:
This innovative approach to fruit segregation revolutionizes traditional grading methods by
introducing automation and machine learning algorithms. It offers enhanced accuracy,
efficiency, and scalability, addressing the challenges faced by conventional fruit handling
processes.

Patent Claims:
1. System Description:
The patent claims describe a system designed to segregate fruits based on their
quality and size.
The system comprises several key components, including conveyor units, image
sensors, bifurcation units, containers, motors, and a controller device.
It is intended to automate the process of fruit segregation, enhancing efficiency and
reducing manual labor.

2. Conveyor Unit:
The conveyor unit serves as the main component for transporting received fruits.
It consists of a first end for fruit intake, a second end, and a body with a rotating belt
to facilitate fruit movement.
A separator element at the second end segregates fruits based on quality criteria.
13
3. Bifurcation Unit:
Positioned at the second end of the conveyor unit, the bifurcation unit further
segregates fruits into different categories.
It includes segregator arms, mid-sections with rotating belts, and partition elements
to divide fruits based on quality and size.

4. Controller Device:
The controller device plays a crucial role in overseeing the operation of the entire
system.
It acquires images of received fruits from image sensors, processes them to
determine quality and size, and controls the motors accordingly.
This ensures efficient segregation of fruits into designated containers.
5. Additional Features:
The patent claims also specify additional features such as sensing modules, fan
assemblies, alert units, and waste disposal mechanisms.
These features contribute to the system's functionality by enabling ambient
parameter monitoring, fan control, alert generation, and waste management.

14
Drawings specification and description:

15
Figure 1 depicts a system architecture (100) designed to segregate fruits based on quality
and size. The system includes components such as a conveyor unit (102), first image sensor
(104), bifurcation unit (106), controller device (128), and others.

● The conveyor unit (102) receives fruits and transports them, while the first image
sensor (104) captures images of the fruits for analysis. A separator element (102-D)
segregates fruits based on quality, controlled by a first motor (102-E).
● The bifurcation unit (106) further segregates fruits based on quality and size using
segregator arms (108, 118), rotating belts, partition elements (108-D, 118-D),
motors (108-E, 118-E), image sensors (110, 120), and division units (112, 122).
● The controller device (128) employs image processing techniques and machine
learning protocols to analyze fruit images, determine quality and size, and control
motor movements for segregation.

Fig. 2 outlines the functional units designed to segregate fruits, comprising the acquisition
unit, first analysis unit, second analysis unit, third analysis unit, and supply control unit, all
operable by the processor of the controller device 128.

● Acquisition Unit (202): Responsible for obtaining images of received fruits, good
quality fruits, and average quality fruits from respective image sensors in real-time
or near real-time via wired or wireless communication.

16
 ● First Analysis Unit (204): Utilizes machine learning protocols such as ANN, CNN,
HMM, AI, or deep learning to analyze acquired fruit images, distinguishing

between first quality (good) and second quality (average) fruits based on pre-stored
databases.
● Second Analysis Unit (206): Focuses on the identified first quality fruits, further
categorizing them into large and small sizes using machine learning techniques like
ANN, CNN, HMM, AI, or deep learning.

● Third Analysis Unit (208): Analyzes the second quality fruits to classify them into
large and small sizes using similar machine learning models as in the second
analysis unit.
● Supply Control Unit (210): Manages the electrical energy supplied to motors
responsible for fruit segregation. It adjusts the position of separator and partition
elements, enabling the segregation process based on fruit quality and size.

Fig. 4 presents tabular data representing the quality assessment of fruits, outlining specific
criteria for categorizing fruit quality on a scale of 1. The key points highlighted in the
illustration are:
17
 ● Quality Scale: Fruits are rated on a scale ranging from 0 to 1, with higher values
indicating better quality.
● Removal of Rotten Fruits: Any rotten fruit with a quality value higher than 0.8 is
identified and removed from the remaining fruits, designated as waste.
● Criteria for Fresh Fruits:
Fruits with a quality level greater than 0.3 and less than 0.6 are categorized as
"Average quality".
Fruits exceeding a quality level of 0.6 are classified as "Good quality".

Fig. 5 presents tabular data showcasing the current ambience parameters of fruits within
reservoirs, specifically the first container 114, second container 116, third container 124, or
fourth container 126. The data is obtained through various sensors and includes the
following parameters:

● Ethylene Level (ppm): Measured using an MQ3 sensor capable of detecting

alcohol concentrations between 25 and 500 parts per million (ppm).

18
 ● Temperature (°C): Recorded using a DHT-11 sensor, with a measurement range of
0 to 50 degrees Celsius and an accuracy of ±2 degrees Celsius.
● Humidity (%RH): Also monitored by the DHT-11 sensor, providing humidity
readings ranging from 20% to 100% with an accuracy of 5%.
● Harmful Gases Detection: Utilizing an MQ135 sensor, capable of detecting
various harmful gases including ammonia, sulfur, benzene, toluene, hydrogen,
carbon dioxide, and smoke, with a detection range from 10 to 1000 parts per million
(ppm).

Fig. 6 showcases an exemplary machine learning protocol designed to segregate fruits

based on their quality, as per embodiments of the present disclosure. The illustration
depicts the following steps:

● Input Image: A fruit image is provided as input to the convolutional neural

network (CNN) model.
● Pixel Analysis: The CNN model analyzes each pixel in the input image,
considering variations in pixel values due to differences in ripeness across the fruit.
● Classification: Based on the pixel analysis, the model determines whether the fruit
is likely to be fresh or ripe.
19
● Threshold Determination: The disclosure employs a threshold value to classify
the fruit as fresh or not. This threshold value is chosen appropriately to optimize
classification accuracy.
● Automatic Threshold Selection: The threshold value can be automatically
determined by constructing a loss function that minimizes error and enhances
evaluation metric scores.

20
Future of process/patent which is patented?

This is a condensed description of how a patentable process or product might develop in the
future:

Market Demand: A patentable process or product's capacity to satisfy market demands

and wishes is a major factor in determining its success. Predicting future demand is aided
by an understanding of consumer preferences and market trends. A patented smartphone
technology, for instance, can have a bright future in the mobile device industry if it
provides cutting-edge features that cater to new consumer demands.

Technological Advancements: Since technology is developing quickly, a patented product

or process's ability to stay ahead of the curve is essential to its success in the future.
Constant innovation as well as Enhancements have the potential to prolong the product's
life and keep it relevant in the marketplace. For example, a manufacturing method that is
patented and integrates automation and machine learning could improve in efficiency and
cost over time, maintaining its competitiveness in the market.

Competitive Landscape: Keeping an eye on rivals and their tactics is crucial to foreseeing
obstacles and possibilities in the future. Finding possible risks and points of differentiation
might be aided by examining rival patents and market positioning. Through persistent
innovation and safeguarding intellectual property, businesses may sustain their competitive
advantage in the marketplace.

Regulatory Environment: A patentable product or process's chances for the future may be
affected by modifications to standards and regulations. To maintain long-term profitability,
21
adherence to safety standards, environmental legislation, and industry-specific needs is
required. adjusting to changes in regulations and taking proactive measures to Concerns

about compliance can reduce risks and improve the product's adoption in the market and
reputation.

Consumer Feedback and Adaptation: Maintaining the success of a patented product or

method requires paying close attention to consumer feedback and making necessary
adjustments in response to shifting preferences. Future growth can be fueled by
incorporating user suggestions, eliminating pain spots, and remaining sensitive to changing
client needs.

22