International Journal on Recent and Innovation Trends in Computing and Communication

ISSN: 2321-8169 Volume: 11 Issue: 10

Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

Importance of Machine Vision Framework with

Nondestructive Approach for Fruit Classification
and Grading: A Review
Dr. D. Baswaraj1, Dr. Mukesh Kumar Tripathi*2, Vajid Khan3, Dr. Neelam Sharma 4, Dr. Satish Addanki5,
Dr. Geetha. K6
1Department
of Computer Science and Engineering,
Vasavi college of Engineering, Hyderabad, India
Email: [email protected]
2Department of Computer Science & Engineering,

Vardhaman College of Engineering, Hyderabad, India

Email: [email protected]
3Department of computer Engineering,

KJ College of Engineering and Management Research, Pune, India

Email: [email protected]
4Department of Physical Education,

Lovely Professional University, India

Email: [email protected]
5Department of Electronics and Communication Engineering,

Mohan Babu University, (Erstwhile Sree Vidhyanikethan Engineering College), Tirupati, India
Email: [email protected]
6Department of Computing Technologies,

SRM Institute of science and Technology, Kattankulathur, India

Email: [email protected]

Abstract— Machine vision technology has gained significant importance in the agricultural industry, particularly in the non-destructive
classification and grading of fruits. This paper presents a comprehensive review of the existing literature, highlighting the crucial role of
machine vision in automating the fruit quality assessment process. The study encompasses various aspects, including image acquisition
techniques, feature extraction methods, and classification algorithms. The analysis reveals the substantial progress made in the field, such as
developing sophisticated hardware and software solutions, which have improved accuracy and efficiency in fruit grading. Furthermore, it
discusses the challenges and limitations, such as dealing with variability in fruit appearance, handling different fruit types, and real-time
processing. The identification of future research needs emphasizes the potential for enhancing machine vision frameworks through the
integration of advanced technologies like deep learning and artificial intelligence.Additionally, it underscores the importance of addressing the
specific needs of different fruit varieties and exploring the applicability of machine vision in real-world scenarios, such as fruit packaging and
logistics. This review underscores the critical role of machine vision in non-destructive fruit classification and grading, with numerous
opportunities for further research and innovation. As the agricultural industry continues to evolve, integrating machine vision technologies will
be instrumental in improving fruit quality assessment, reducing food waste, and enhancing the overall efficiency of fruit processing and
distribution.
Keywords- Non-destructive, Near-infrared, quality grading, machine learning

I. INTRODUCTION become evident after harvest, causing post-harvest losses and

diminishing fruit quality. Traditionally, the evaluation of mango
Mango (Mangifera indica) is one of the most popular and
peel color and pigment content has been performed through
economically significant tropical fruits globally. Among the
destructive methods that involve peeling the fruit or extracting
numerous mango cultivars, 'Keitt' stands out for its exceptional
pigments. These methods are not only time-consuming but also
flavor, nutritional value, and export potential [1]. However, the
result in fruit wastage, making them less practical for large-
occurrence of internal physiological disorders, such as chilling
scale quality assessment in the industry. Non- destructive
injury, internal breakdown, and flesh browning, poses a
technologies, particularly spectroscopy and imaging, present a
significant challenge to both growers and consumers. These
promising alternative. Spectroscopy entails the exposure of
disorders are often hidden from external inspection and only
mango peel to light across the electromagnetic spectrum,

970
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

capturing the wavelengths of light absorbed, transmitted, or solely based on external features. Therefore, developing
reflected by the peel. The resulting spectral data can yield effective non-destructive techniques for disease and defect
valuable insights into the color and pigment composition of the detection in mangoes is essential to ensure the quality and
peel without physically altering the fruit. This study unites non- marketability of these fruits while streamlining the sorting and
destructive techniques such as spectroscopy and imaging with handling processes.
advanced data analysis and modeling, leading to the
II. INTERNAL QUALITY PARAMETERS FOR MANGO
development of reliable and swift methods for predicting color
and pigment contents in mango peel. The outcomes of this Consumer preferences for agricultural products are influenced
research could revolutionize mango quality assessment by various factors, with quality and cost being primary
practices, curbing waste and ensuring the steady delivery of considerations. The internal attributes mentioned, such as Total
high-quality mangoes to growers, processors, distributors, and Acidity (TA), Dryness, Firmness, Maturity, and Un-maturity,
consumers. As the global demand for fresh, nutritious produce play a crucial role in determining the overall quality of
continues to surge, the capability to non-destructively forecast agricultural products, including mangoes. Let's briefly discuss
color and pigment contents in mango peel promises to make a these attributes: Total Acidity (TA), Total Acidity (TA),
substantial impact on the success and competitiveness of the Firmness, and Maturity. To address this challenge, consumers
mango industry. can learn to select mangoes based on visual and tactile cues,
The recognition of color is a crucial parameter for such as color and firmness. Additionally, sellers can provide
evaluating the external quality of mangoes. It involves assessing information about the expected ripening time for the mangoes
the appearance and surface of mango fruits based on their skin they sell, helping consumers make more informed choices [4].
or outer surface. Visual inspection is a straightforward method, In summary, understanding the internal attributes of agricultural
with variations mainly depending on changes in lighting and products like mangoes is essential for consumers and sellers to
illumination. However, traditional color measurement methods make informed decisions about quality and ripeness. Still, the
have limitations, even though basic color information can be dynamic nature of these attributes due to ripening makes it
directly obtained from image datasets. Standard color necessary to develop strategies for better selection and
measurement typically assumes a uniform and consistent communication between sellers and consumers.
surface. When the surface of an object has multiple colors or is The ripening process of mango fruits is indeed
non-uniform, measures must be repeated several times to influenced by temperature, which affects their physicochemical
capture the color variations accurately. In cases where the attributes over time. Temperature plays a crucial role in the
surface area of the thing, such as a mango, exhibits diverse ripening process of mango fruits. Lower temperatures, like
colors, estimating the color becomes a challenging task, 20°C, are better for maintaining fruit quality over an extended
especially when dealing with a large volume of fruits with non- period, while higher temperatures, such as 25°C, might be
homogeneous characteristics [2]. As a result, there is a pressing suitable for quickly ripening mangoes with sweeter and more
need to develop rapid and non-destructive methods for accurate acidic characteristics. Still, they may not have the same shelf
color estimation, particularly for assessing the color distribution life as those matured at lower temperatures. Physiological loss
in agricultural products. Such advancements in color in weight is a measure of moisture loss during ripening [5]. It is
measurement techniques are crucial for improving the precision mentioned that at 20°C, mango fruits experience less
and efficiency of evaluating the external quality of fruits like physiological loss in weight compared to 25°C. This indicates
mangoes in agricultural production. that the fruit retains more moisture at the lower temperature.
Disease assessment is another critical parameter in The information you provided summarizes the biochemical
evaluating the external quality of mango fruits. Common parameters of mango fruits, including their moisture content,
diseases that affect mango fruits include over-ripeness and carbohydrate content, total soluble solids (TSS), total sugar
bruises. While mango deterioration is inevitable, non- content, and the levels of vitamins A, B, C, and nicotinic acid
destructive methods provide a solution for assessing external [6]. It also mentions the change in sugar content as mangoes
characteristics to detect various types of defects during the ripen and the similarity in moisture content between unripe and
handling and processing of mangoes. Identifying superficial ripe mangoes. The author [7] shows in their research that a sugar
weaknesses is vital, as it can significantly impact the quality and component shows a symmetric increment at a mature state.
market value of mango fruits [3]. However, grading mangoes Initially, the sugar value is less at the unripe stage, but the sugar
based on their external parameters presents challenges due to component was found to be gained in maturity. The range of
several factors. Mango fruits exhibit variations in illumination, moisture in ripe mango is very similar to unripe mango, with a
shape, and ripeness conditions. These variations make it value of 78.9 % to 82.8 % per 100 g.
difficult to establish a consistent and accurate grading system

971
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

The information provided from the author's [8] affect the eating experience.
discussion highlights the variation in bio-accessible β-carotene Juiciness: Juiciness is an important quality parameter. Ripe
content in different mango varieties, emphasizing the mangoes should be juicy, but not excessively so.
importance of this antioxidant compound in mangoes. In Astringency: Some mango varieties may have astringent
addition to β-carotene content, the author also mentions various properties when unripe. As they ripen, this astringency
parameters that can affect the quality of mangoes during the decreases, which is desirable for consumption.
ripening stage. These include temperature, controlled and Taste and Flavour: The overall taste and flavour profile,
modified atmosphere storage, ionizing radiation, and using including hints of tropical and citrus notes, contribute to the
chemicals for ripening or preservation. These factors play a mango's desirability.
significant role in determining the shelf life and overall quality Defects: Any internal defects, such as bruising or rot,
of mangoes. Furthermore, the author suggests that internal negatively impact the quality of the mango.
parameters such as TAC (Total Antioxidant Capacity), SSC
III. PHYSICAL PARAMETERS FOR RIPPED MANGO
(Soluble Solids Content), and TA (Titratable Acidity) are
crucial for grading mango quality. Monitoring these internal The width of most mango varieties is relatively consistent, with
parameters is essential as relying solely on visual appearance slight variation. 'Mallika' and 'Alphonso' varieties have the
may not be sufficient to detect internal defects or variations in largest width. 'Samarbahist Chausa' has a narrower width of
quality. Early recognition of internal features is necessary, as 57.0 mm compared to the other varieties. The dimensions of
the severity of defects can impact the financial value of mangoes, such as length and width, are essential physical
mangoes. To address this, non-destructive methods are characteristics that can influence their appearance and, to some
recommended to identify internal characteristics, reduce waste, extent, their eating quality. These parameters help categorize
and maintain the overall quality of mangoes throughout the and differentiate mango varieties based on their shape and size.
supply chain. These internal quality parameters are often Additionally, specific varieties, such as 'Totapuri,' stand out
assessed visually, through taste tests, and by using various tools with distinctive characteristics, in this case, having the most
and instruments to measure specific attributes like sugar content extended length. The width of most mango varieties is relatively
and acidity. Accurate evaluation of these parameters is essential consistent, with slight variation. 'Mallika' and 'Alphonso'
for ensuring that mangoes are harvested, stored, and distributed varieties have the largest width. 'Samarbahist Chausa has a
at the peak of their quality, which is crucial for domestic and narrower width of 57.0 mm compared to the other varieties [7].
international mango markets. Mangoes, like many fruits, are The information provided highlights various physical
assessed for their internal quality based on various parameters parameters and characteristics of different mango varieties,
that determine their taste, texture, ripeness, and overall including weight, volume, and specific gravity, as well as
desirability. Here are some of the critical internal quality carotenoid content. Mallika mangoes have the highest width,
parameters for mangoes: measuring 382.23 g. They also have the highest importance,
Sugar Content (Brix): The sugar content, typically measured with a value of 373.33. Fazli and 'Langra' Varieties follow
in degrees Brix, indicates the sweetness of the mango. Ripe 'Mallika' regarding width and volume. ‘Samarbahist Causa’s
mangoes have higher sugar content, making them sweeter. mangoes are comparatively lighter, weighing 173.33 g. They
Acidity (pH): The acidity level influences the mango's taste. also have a smaller volume, measuring 93.7 ml. Specific gravity
The right balance of sweetness and acidity is important for a is a measure of density, and it varies among different mango
pleasant flavor. The pH value can also be an indicator of varieties, ranging from 1.095 to 1.022. This indicates
ripeness. differences in thickness and composition among the mango
Aroma: The aroma or fragrance of a mango is a crucial quality varieties.' Alphonso and 'Fazli varieties have the lowest
parameter. A ripe mango should have a distinct, pleasant aroma carotenoid content, with values of 2.65 mg/100 and 1.82
that reflects its variety. mg/100, respectively. Carotenoids are pigments responsible for
Firmness: The firmness or texture of the mango is assessed by the color of many fruits and vegetables. These parameters,
touch. It should be neither too hard nor too soft. The texture including weight, volume, specific gravity, and carotenoid
changes as the mango ripens. content, provide valuable information for assessing and
Flesh Colour: The colour of the mango flesh can vary by characterizing different mango varieties. The variations in these
variety. It can be yellow, orange, or even red. The internal attributes can influence the fruit's appearance, taste, and
colour is an indicator of ripeness, with most mangoes turning nutritional value. Consumers and the industry must understand
from green to a more vibrant colour when ripe. these differences to choose the mango variety that suits their
Fiber Content: Some mango varieties have fibrous flesh, while preferences and needs.
others are fibre-free. The presence and quantity of fibres can

972
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

The presence and levels of carotenoid content in Skin Texture: The texture of the mango skin can be smooth or
mango varieties can vary, and these variations are often due to slightly wrinkled, depending on the variety. It should be firm
specific carotenoids, including β-carotene, lycopene, lutein, and but yield to gentle pressure when ripe.
zeaxanthin. Carotenoids are responsible for the coloration of Flesh Texture: The flesh of ripe mangoes should be smooth
many fruits and vegetables and are crucial for their nutritional and juicy. The texture can vary from buttery to fibrous,
value. Identifying fruit ripening often depends on measurable depending on the variety. Some mangoes are known for their
attributes such as titratable acidity [8]. Titratable acidity is a fibre-free flesh.
measure of the fruit's acidity, and it can change with the ripening Aroma: Ripe mangoes are known for their pleasant, fruity
process. Ripening is also influenced by prevailing fragrance. The intensity and type of aroma can differ among
environmental conditions, including temperature and humidity, mango cultivars.
as well as ecological factors. The variations in acidity levels Fruit Maturity Stage: It's essential to assess the mango's
among different mango varieties can be attributed to their maturity stage. A ripened mango should be neither underripe
inherent characteristics, which are influenced by environmental nor overripe. The set of ripeness can affect the texture and taste.
and ecological conditions. These variations are natural and can Seed Size: The size and shape of the seed or stone inside the
be considered a part of the mango's genetic diversity. The mango can vary among cultivars. Some varieties have smaller
increase in ascorbic acid (vitamin C) content in mangoes during roots, which means more edible flesh.
their growth and development may be due to the continuous Juiciness: Ripe mangoes should be juicy, and the juice content
synthesis of glucose 6-phosphate, a precursor for ascorbic acid. can vary among different mango cultivars.
This increase may also result from the catalytic effect of Taste and Flavour: The taste and flavour of mangoes vary
growth-promoting substances on the biosynthesis of ascorbic widely between cultivars. Some are exceptionally sweet and
acid from sugars [9]. In summary, the presence of carotenoids, have unique flavour profiles.
the changes in titratable acidity, and the increase in ascorbic
IV. NON DESTRUCTIVE MEASUREMENT
acid content in mangoes are complex processes influenced by
genetic factors, environmental conditions, and biochemical A non-destructive estimation is the assessment of materials in
pathways. These factors contribute to the unique qualities and terms of shape, biochemical, and qualities without crushing
nutritional profiles of different mango varieties. their usefulness, just as the framework highlights. The
The physical parameters of ripened mango cultivars conventional approach to identifying the inner attributes of fruit
grown in India can vary depending on the specific variety, needs a destructive technique using apparatus, strategies, and
growing conditions, and maturity stage. These physical training professionally, which brings about high examination
parameters are essential for grading, quality assessment, and costs. All this author says that [10] new advancements to keep
marketing mangoes in India. The particular values and track of fruit quality changes during the post-harvest taking care
characteristics of these parameters can differ from one mango of chain are quickly being presented, particularly those
cultivar to another, and the mango industry often relies on dependent on non-destructive evaluation strategies. Hence,
grading standards to ensure consistency and quality in the non-destructive methods allow supplanting the standard
marketplace. However, here are some of the standard physical reference strategy, including the traditional methodology. Right
parameters used to assess ripened mangoes: now, non-destructive methods are utilized in mechanical
Size: The size of ripened mangoes can vary significantly applications to guarantee the practicability of a given item or
depending on the cultivar. Common mango varieties in India framework. Non-destructive methods are significant in deciding
include Alphonso, Kesar, Langra, and Dasheri, and each has its the dependability of different applications, remarkably in
characteristic size and shape. assessing the quality parameters of fruits. These efficient and
Weight: Mangoes can vary in weight based on their size and non-ruinous strategies can benefit by furnishing definitive
maturity. The weight is an essential parameter for grading and attributes to acquire better quality mango items and advancing
pricing. utilization of mangoes with better well-being benefits.
Shape: Mangoes can be oval, round, or somewhat oblong, and A. Infrared based spectroscopy
their shape is a distinguishing feature of different varieties.
The author [11] has used VIS-NIR spectroscopy with a robot
Colour: The external colour of ripened mangoes varies with the
gripper for mango quality grading. They have also presented the
variety. For instance, Alphonso mangoes are known for their
framework which describes the working model of spectroscopy
bright yellow or orange skin, while Langra mangoes often have
in robot gripper. The author [12] has also utilized the near-
a greenish-yellow skin when ripe.
infrared reflection spectroscopy to acquire the spectra. The
range of NIR spectroscopy is between 400 to 700 nm. The
passage describes a study using VIS-NIR spectroscopy with a

973
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

robot gripper for mango quality grading. Here are the key points technique, a pattern estimation of value changes during fruit
from the selection: product improvement dependent on inspecting different mango
Use of VIS-NIR spectroscopy: The author used visible and samples is never again required. Some author, for example,
near-infrared (VIS-NIR) spectroscopy to assess mango quality. [15], have enclosed in their research the full range utilizing
Framework for spectroscopy in robot gripper: The author spectroscopy in mangoes, even though reviews using colour
presented a framework that outlines how spectroscopy can be arranges are increasingly expected, for example, [16].
integrated into a robot gripper for mango quality grading. This Over the past few years, X-ray has given a quality-
suggests an automated system for assessing mango quality. based 3D image of objects with various materials, making a
Near-infrared reflection spectroscopy: Near-infrared (NIR) difference in separating low-and high-density materials. To
reflection spectroscopy was employed to capture the spectra of fulfil this intention, the author [17] has studied the physical
the mango samples. The NIR spectra were obtained in the range properties of cucumber fruit. The physical properties such as
between 400 to 700 nm, with readings taken at 10 nm intervals. density, porosity, moisture content, and Firmness is estimated
Data collection: Mango data samples were collected for for seven continuous days at different temperature.
experimental purposes. These samples were kept at a specific Respectively, X-ray techniques are applied to the sample to
temperature (24±10°C) before the measurement of the spectra. collect the 3D –image. Further segmentation method was
Temperature control is important to ensure consistent performed to obtain the average gray value of the pixel. Finally,
measurements. by using regression techniques, the change in fruits is analysed.
Raw spectra and noise removal: The raw spectra collected They have also expressed that in higher temperatures of 25 C,
had high dimensions. To prepare the data for analysis, a pre- the average value grayscale decreases, whereas increment is
processing step was conducted to remove noise from the raw observed in the internal porosity parameter. In their study,
spectra. This was achieved by applying weight regression authors [18] evaluated the porosity in fruits and vegetables by
coefficients. applying X-ray CT. They also presented the framework where
Analysis of internal parameters: After pre-processing, an X-ray machine scanned an image sample. This machine has
internal parameters like Total Soluble Solids (TSS), Total ranged between 746 mm to 1085 mm with 120 KV. Further, two
Acidity (TA), and firmness were analysed using a Partial Least tissue sample with high and low resolution is analysed.
Squares (PLS) model. PLS is a technique used for diagnosing
B. Machine vision
the relationship between predictor variables (spectra) and
response variables (in this case, TSS, TA, and firmness) to make Machine vision is a part of an engineering expert skill in
predictions or assessments. consolidation with optical and mechanical properties,
The study appears to be focused on developing a system for electromagnetic, and image processing methods. Machine
automated mango quality grading using spectroscopic data, and vision frameworks are progressively faster regarding sample
the passage describes the data collection, pre-processing, and image assessment when analyzing acoustic and vibration
analysis steps involved in the research. The author presents a techniques. The improvement of the application has increased a
framework based on electrical impedance spectroscopy to lot of enthusiasm for the quality assessment of agricultural
identify maturity levels to assess mango quality [13]. In this, products. The volume estimation is essential for fruit to decide
‘Tommy Atkins’ variety of mango has been utilized in the the size per specific grading evaluations [19]. At the present
experiment. The data sample was washed and kept at different stage, the principle worries in this machine vision explore the
degrees of temperature for 15 days before the experiment. The capacity of machine vision to classify the agriculture product
electro-mechanic universal machine with TESC software is regarding quality parameters. Till now, research has dealt with
utilized for mechanical tests. With this, the mango sample was the primary practice of quality assessment with methods for
connected by two electro nodes. Finally, the maturity of mango machine vision frameworks. But, just a couple of the present
can be obtained by calculating fitting results divided by mango works discussed the physicochemical and biochemical presence
fruit diameter. The author [14] presents a NIR spectral in the image sample [20].
framework with a combination of an NIR spectrometer and a In a computer vision-based system, each image is
tool for mango fruit positioned, in this, internal parameters such represented in some particular color space. Many color spaces
as Dry matter (DM), TA, SSC, and ripeness will be reviewed to are available in image processing. Some common color space is
grade the quality of mango fruit. At last, MLR and PLS RGB color space, HSI color space, L* a* b* color space, etc.
regression model is used to analyse the data. The outcomes [21]. For example, the RGB color space is frequently utilized in
show that NIR is a promising strategy to explore the exacting R G B individually of mango fruits, which contains the
relationships of biochemical parameters of the same mango three wavelengths with the composition of red, green, and Blue.
product at various formative stages. With this recently created As we know, RGB color space is hardware-based. Color

974
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

changes have been done by the standard color value of the calculated by the first Quadrant method. They have utilized the
particular image that can be identical to a human person in HSI L* a* b* color space. Finally, in the calibration model, both
color space [22]. The machine vision framework plays a vital mango 'Nam Dokmai' and 'Maha Chanok' categories are
role in detecting color, texture, shape, and illumination. classified by static methods such as Mean and standard
Identifying disease or external damage is ongoing with further, derivation. The grading of mangoes is done based on three
more challenging tasks. Particularly since the present different types: ripe, unripe, and overripe. The author presents
measurement relies upon the high discrepancy of disease. For an artificial intelligence-based structure for sorting mangoes
instance, the exactness of the machine vision framework to [29]. Their study briefly described the role of image processing
assess the external part of food products depends upon a few and AI in sorting the mango. The present model helps capture
elements, including the cultivar, planting area, and postharvest the image of the mango. Next, the segmented process and all
treatment of fruits [23]. features are extracted to obtain a segmented image of the mango
The author [24] has presented a computer vision based on color, mass, and volume features. Finally, ANN-based
approach to improve productivity. This system comprises a regression method quality grading has been done. The proposed
camera, frame grabber, and sample holder component. In this, system has produced 80 percent accuracy rates for sorting
image samples are kept on a sample holder. This image holder mangoes. A hardware and software-based model system is
is connected to the camera, which captures the image and explored to sort a mango by the author [30-32]. They have
forwards it to the next stage, where the image is collected in utilized a conveyor belt, AC motor controller, webcam, sensor,
frame format using the frame grabber component. The system and serial combination board. In this model, mango samples are
is connected by wired with all members of the proposed kept on conveyor belts that pass through a webcam device to
framework. They have analyzed the review of various fruits and capture the image. Then, the image is segmented by Threshold
vegetables. Image processing techniques mostly do the quality techniques. The dilatation and erosion method is applied to the
analysis of fruits and vegetables. This technique consists mainly respective segmented image to remove the unwanted noise or
of four steps: the first step is to do pre-processing to remove shadow. This dilatation and erosion operation plays a vital role
unwanted noise, the second segmentation process to extract the in improving the performance of the proposed system. Further,
background of the image sample, feature extraction to extract the size of the mango sample is based on the caliber
the parameter responsible for the quality, and finally, applying determination method. This method extracts the pixel value of
classifiers to detect the proposed system's performance. images based on the weight of the respective image. The
The author [25] first prepared the image acquisition proposed system achieves a percentage of accuracy rate. The
chambers to grade the mango quality. A design framework was quality of mango fruit significantly relies on numerous
presented by the author [26]. The layout of the system is helpful parameters, such as ripe, unripe, overripe, texture, shape, color,
for fruit grading. This module combines components such as and other factors at harvesting time.
conveyor belts, fruit placers, image-capturing chambers, fruit Object recognition and classification is an essential
sorting strips, and sorting bins. In this back propagation neural application in computer vision systems. The framework
network (BPNN) is utilized at the classification stage to sort the presented by the author [33] consists of four parts. Firstly, the
fruit based on respective quality. 80.0 percent accuracy rate is mango image is segmented Ostu Threshold- Techniques. Next,
achieved by the proposed system for grade II fruit, whereas 71 the feature is extracted by using Region and Boundary
percent accuracy for grade I and 66 percent for grade III date descriptors. Finally, a 90.01 percent accuracy rate is achieved
fruit has been achieved. The author presents a computer vision by applying Bayes classifier. A similar type of framework is
framework for mango grading [27]. Author [28] also presents presented by the author [34]. They must grade the mango by
the Machine Vision system for the quality determination of regression model based on mass and volume features. The
mango. This framework consists of components such as proposed system produces a 91.76 percent accuracy rate for
Conveyed assembly, Electric power drive, Fruit samples, grading mangoes.
Illumination unit, Light sources, Camera, Control unit,
Computer, Frame grabber software, and Variable-frequency V. ADVANTAGES OF NON-DESTRUCTIVE METHOD
control. For the experimental purpose, the author collected two Non-destructive methods offer several advantages in various
different categories of mango, 'Nam Dokmai' and 'Maha fields, including agriculture, food processing, quality control,
Chanok'. and scientific research. Overall, non-destructive methods are
Further, all sample images of mangoes were versatile and essential in many applications, offering a balance
segmented by the Gamma Curve Fitting method. All the between thorough assessment and preservation of samples,
backgrounds extracted from the mango image are processed in efficiency, and cost-effectiveness. They play a crucial role in
the next step, where the color of each mango sample is quality control, research, and safety across various industries.

975
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

Here are some of the key advantages of non-destructive VI. DISDVANTAGES OF NON-DESTRUCTIVE METHOD
methods:
While non-destructive methods offer many advantages, they
➢ Preservation of the Sample: Non-destructive methods
also have certain disadvantages and limitations. It's important
allow for the evaluation and assessment of a sample
to consider these drawbacks when choosing between non-
without causing any damage or alteration. This is
destructive and destructive methods for specific applications.
particularly valuable when dealing with limited or valuable
It's essential to carefully evaluate the specific requirements of
specimens.
an application and weigh these disadvantages against the
➢ Reduced Waste: Since non-destructive methods do not
benefits of non-destructive methods when choosing the
damage or consume the sample, there is minimal to no
appropriate testing or measurement approach. In some cases, a
waste generated during the testing or analysis process.
combination of non-destructive and destructive methods may
➢ Cost-Effective: Non-destructive methods can be cost-
provide the most comprehensive assessment. Here are some of
effective in the long run. While the initial investment in
the disadvantages of non-destructive methods:
equipment and technology may be significant, it often leads
➢ Limited Sensitivity: Non-destructive methods may have
to savings by reducing the need for extensive sample
limitations in their ability to detect subtle changes or
preparation or disposal costs.
defects, especially in cases where the changes are below the
➢ Efficiency: Non-destructive methods are often faster and
method's detection threshold. This can lead to false
more efficient than destructive ones. They can provide
negatives or missed issues.
immediate results and reduce the need for time-consuming
➢ Cost of Equipment: The initial cost of acquiring and
sample preparation or retesting.
maintaining non-destructive testing or measurement
➢ Safety: Non-destructive methods are generally safer for
equipment can be high, making it a barrier for smaller
operators since they do not involve the handling of
organizations or those with limited budgets.
hazardous or toxic materials or the risk of exposure to
➢ Complexity: Some non-destructive methods can be
harmful conditions.
complex to implement and require skilled operators or
➢ Non-Invasive: Non-destructive methods are non-invasive,
technicians to obtain accurate results. This complexity can
making them suitable for sensitive or fragile materials,
lead to higher operational costs and potential for human
living organisms, and cultural artifacts. They are
error.
commonly used in medical diagnostics, archaeological
➢ Sampling Variability: The accuracy of non-destructive
studies, and the assessment of historic objects.
methods can be affected by variability in sample properties,
➢ Real-Time Monitoring: Many non-destructive methods,
shape, and orientation. Variations in sample composition
such as real-time imaging and sensor technologies, provide
can also impact the results.
continuous or frequent monitoring of processes, structures,
➢ Calibration and Standardization: Non-destructive
or materials, which is essential for quality control and
methods often require calibration and standardization to
safety.
provide reliable results. Failing to maintain the equipment
➢ Multiple Parameters: Non-destructive methods can
and standards can lead to inaccuracies.
simultaneously assess multiple parameters or properties of
➢ Time-Consuming: Non-destructive testing can sometimes
a sample, providing a comprehensive view of its
be time-consuming, particularly when it involves extensive
characteristics. For example, near-infrared spectroscopy
setup, data collection, and analysis. This may not be
can assess various chemical properties of a food product in
practical in situations that require rapid results.
a single measurement.
➢ Sustainability: Non-destructive methods align with VII. MACHINE LEARNING IN FOOD INDUSTRIES
sustainability principles by reducing waste and resource
Machine learning is increasingly being applied in the food
consumption. They support environmentally friendly
industry to enhance various aspects of food production,
practices by minimizing the impact of testing processes.
processing, quality control, and distribution. Here are some key
➢ Quality Control: In industrial settings, non-destructive
areas where machine learning is making a significant impact in
methods are crucial for quality control and ensuring that
the food sector: The author [36-38] has introduced a framework
products meet the required specifications without
by NIR spectroscopy. In their study, non-infrared spectroscopy
compromising the integrity of the samples.
is utilized to assess the internal features of mangoes. Machine
➢ Non-Contamination: Non-destructive methods eliminate
learning in the food industry relies on a wealth of data sources,
the risk of contamination, ensuring that the sample remains
including sensor data, image data, chemical analyses, and
in its original state and free from contamination that could
consumer preferences. By leveraging this data, machine
occur during handling or analysis.
learning models can improve the efficiency and quality of food

976
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

production and distribution while addressing critical issues like which indirectly impacts the quality and availability of
food safety and waste reduction. food products.
Quality Control and Inspection: Waste Reduction:
➢ Machine learning algorithms can be used to automate the ➢ Machine learning systems can help reduce food waste by
inspection of food products for defects, contamination, and optimizing production, distribution, and inventory
quality issues. Computer vision systems can identify visual management.
anomalies and sort out subpar products. ➢ Apps and platforms can provide consumers with
Food Safety and Traceability: information on minimizing food waste through recipes and
➢ Machine learning can help in food safety by predicting and meal planning.
identifying potential contaminants, pathogens, and
VIII. PROBLEM IDENTIFICATION
allergens. It can also be used for early warning systems to
detect outbreaks. Grading the quality of mango fruit is a crucial process in the
➢ Machine learning are employed to track and trace food fruit industry to ensure that only mangoes of desired quality
products throughout the supply chain, ensuring reach the market or processing facilities. However, this task can
transparency and authenticity. be challenging due to the variability in mango quality and the
Quality Prediction and Shelf-Life Estimation: subjectivity involved in human grading. Using automated
➢ Machine learning models can predict the quality and shelf systems and technology, such as machine learning and
life of food products based on various factors, including computer vision, can help address these issues. Grading the
storage conditions, packaging, and initial quality. quality of mango fruit is currently a manual and subjective
➢ These predictions are valuable for inventory management process, which can lead to inconsistencies and inefficiencies in
and reducing food waste. the fruit supply chain. There is a need for an automated,
Supply Chain Management: objective, and accurate grading system that can assess various
➢ Machine learning is used for demand forecasting, inventory quality parameters, including color, size, ripeness, and external
optimization, and logistics planning. It helps in reducing defects, to ensure uniform, high-quality mangoes are delivered
food spoilage and optimizing distribution. to consumers, processors, and export markets. This system
➢ Predictive analytics can also mitigate the impact of supply should also accommodate different mango varieties, as their
chain disruptions or unexpected events. quality parameters may vary.
Food Authentication and Fraud Detection:
IX. PROPOSED ARCHITECTURE
➢ Machine learning can be employed to detect food fraud and
adulteration by analyzing various data sources, such as While non-destructive techniques offer several advantages, it's
chemical compositions and spectral data. important to acknowledge their limitations, including the initial
➢ It can authenticate the origin and quality of products, cost of technology acquisition, the need for calibration and
including verifying the authenticity of premium labels like standardization, and potential challenges related to
organic or certified products. environmental conditions and sample variability. The
Nutritional Analysis: development and adoption of non-destructive grading methods
➢ Machine learning models can estimate the nutritional in the mango industry hold great promise for enhancing the
content of food products based on their ingredients and efficiency and reliability of mango quality assessment, ensuring
composition. consumers receive high-quality, consistent, and safe fruit while
➢ Apps and platforms use machine learning to provide also benefiting the industry by reducing operational costs and
nutritional recommendations and personalized dietary minimizing waste. Further research and collaboration between
advice to consumers. technology providers, mango producers, and regulatory bodies
Flavor Profiling and Development: will be crucial in the successful implementation and
➢ Machine learning can help food companies create new standardization of non-destructive grading systems for mango
flavor profiles by analyzing ingredient combinations, fruit. Creating an architecture to grade the quality of mango fruit
aroma compounds, and sensory data. involves integrating various technologies and components to
➢ It can assist in developing unique and appealing food automate the grading process efficiently and accurately. This
products. architecture provides a high-level overview of the features and
Precision Agriculture: stages of a non-destructive mango grading system. The
➢ In the context of agriculture, machine learning is used for successful implementation of such a system requires a
crop management, irrigation control, and disease detection, collaborative effort between technology providers, mango
producers, and regulatory bodies to establish industry standards

977
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

and ensure the quality and consistency of mango grading • Regular Maintenance: Establish a maintenance schedule
processes. Below is a high-level architecture that outlines the to ensure the continued accuracy and reliability of sensors
key components and stages of a non-destructive grading system and equipment.
for mangoes: • Calibration: Regularly calibrate the system to account for
1. Data Acquisition: Utilize a combination of sensors, changes in environmental conditions and ensure the quality
cameras, and imaging systems, including hyperspectral assessments remain accurate.
cameras, near-infrared (NIR) spectroscopy, and colour
cameras, to capture data from mangoes. XI . CONCLUSION
2. Data Preprocessing: Data Cleaning and Enhancement: In conclusion, non-destructive techniques for grading the
Apply preprocessing techniques to clean and enhance the quality of mango fruit offer a promising approach to enhance
raw data, including noise reduction, outlier removal, and the efficiency and accuracy of mango quality assessment. This
calibration. review highlights the key points and advantages of using non-
3. Feature Extraction: destructive methods for mango grading. In this paper, we
Extract relevant features from the data, such as colour, size, surveyed fruit's internal and external parameters to grade the
spectral signatures, and texture, for the grading process. quality based on their properties. We additionally analyze the
4. Real-Time Grading: non-destructive measurement to evaluate the internal
• Integration with Conveyor Systems: Implement the parameter; further, various method of pre-processing spectra is
grading system in conveyor lines to allow mangoes to pass discussed. Finally, the role of machine learning techniques in
through the inspection area. the food industry is briefly examined. Literature review and
• Computer Vision Cameras: Use high-resolution cameras analytical comparison have been made among recent state-of-
and sensors to capture images and data from mangoes in the-art internal and external features applying destructive and
real time. destructive, based on which the five most challenging research
• Quality Prediction: Apply the machine learning models to gaps are identified that need to be addressed.
predict the quality of each mango as it passes through the
REFERENCES
inspection area.
5. Grading and Sorting: [1]C. S. Nandi, B. Tudu, and C. Koley, “A machine vision technique
for grading of harvested mangoes based on maturity and
• Quality Grading: Based on the predictions, assign a
quality,” IEEE Sens. J., vol. 16, no. 16, 2016, doi:
quality grade to each mango, such as "premium," 10.1109/JSEN.2016.2580221.
"standard," or "reject." [2]V. E. Nambi, K. Thangavel, and D. M. Jesudas, “Scientia
• Sorting Mechanisms: Implement sorting mechanisms, Horticulture Scientific classification of ripening period and
such as robotic arms or air jets, to divert mangoes into development of colour grade chart for Indian mangoes (
appropriate bins or conveyor lanes based on their quality Mangifera indica L .) using multivariate cluster analysis,” Sci.
grade. Hortic. (Amsterdam)., vol. 193, pp. 90–98, 2015.
6. Data Storage and Analytics: [3] Tripathi, M.K. and Maktedar, D.D., 2018. A Framework with
OTSU’S Thresholding Method for Fruits and Vegetables Image
• Data Logging: Store all grading data, including quality
Segmentation. International Journal of Computer Applications,
assessments and environmental conditions, for traceability 975, p.8887.
and analysis. [4] Tripathi, M.K. and Maktedar, D.D., 2016, August. Recent machine
• Data Analytics: Analyse historical data to identify trends, learning based approaches for disease detection and
improve quality, and optimize the grading process. classification of agricultural products. In 2016 International
7. Reporting and User Interface: Conference on Computing Communication Control and
• User Interface: Develop a user-friendly interface for automation (ICCUBEA) (pp. 1-6). IEEE.
[5]Y. Ding, L. Wang, Y. Li, and D. Li, “Model predictive control and
operators and quality control personnel to monitor the
its application in agriculture: A review,” Computers and
grading process, adjust settings, and view quality reports.
Electronics in Agriculture, vol. 151. 2018, doi:
• Quality Reports: Generate detailed quality reports for 10.1016/j.compag.2018.06.004.
each batch of mangoes, including statistics on the [6]S. C. N. Aleixos and F. A. A. Torregrosa, “Optimized computer
distribution of quality grades. vision system for automatic pre-grading of citrus fruit in the field
8. Integration with Supply Chain: Connect the grading using a mobile platform,” no. 45, 2013, doi: 10.1007/s11119-
system with the broader supply chain, allowing for seamless 013-9324-7.
tracking and traceability of mangoes from the point of grading [7]A. K. Bhatt and D. Pant, “Automatic apple grading model
to the consumer. development based on back propagation neural network and
machine vision , and its performance evaluation,” 2013.
9. Maintenance and Calibration:

978
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

[8]M. Shamili, “The estimation of mango fruit total soluble solids visible imaging,” Comput. Electron. Agric., vol. 115, pp. 51–56,
using image processing technique,” Sci. Hortic. (Amsterdam)., 2015, doi: 10.1016/j.compag.2015.05.006.
vol. 249, no. October 2018, pp. 383–389, 2019, doi: [22] Tripathi, Mukesh Kumar, and Dhananjay D. Maktedar. "Internal
10.1016/j.scienta.2019.02.013. quality assessment of mango fruit: an automated grading system
[9]V. Mohammadi, K. Kheiralipour, and M. Ghasemi-varnamkhasti, with ensemble classifier." The Imaging Science Journal 70, no.
“Detecting maturity of persimmon fruit based on image 4 (2022): 253-272.
processing technique,” Sci. Hortic. (Amsterdam)., vol. 184, pp. [23]K. Utai, M. Nagle, S. Hämmerle, W. Spreer, B. Mahayothee, and
123–128, 2015, doi: 10.1016/j.scienta.2014.12.037. J. Müller, “Mass estimation of mango fruits (Mangifera indica
[10]A. Mohapatra, S. Shanmugasundaram, and R. Malmathanraj, L., cv. ‘Nam Dokmai’) by linking image processing and artificial
“Grading of ripening stages of red banana using dielectric neural network,” Eng. Agric. Environ. Food, vol. 12, no. 1, pp.
properties changes and image processing approach,” Comput. 103– 110, 2019, doi: 10.1016/j.eaef.2018.10.003.
Electron. Agric., vol. 143, no. 382, pp. 100–110, 2017, doi: [24] Chiranjeevi, Kasa, Mukesh Kumar Tripathi, and Dhananjay D.
10.1016/j.compag.2017.10.010. Maktedar. "Block chain technology in agriculture product supply
[11]X. Chen, Z. Li, Y. Wang, and J. Liu, “Effect of fruit and hand chain." In 2021 International Conference on Artificial
characteristics on thumb – index fi nger power-grasp stability Intelligence and Smart Systems (ICAIS), pp. 1325-1329. IEEE,
during manual fruit sorting,” Comput. Electron. Agric., vol. 157, 2021.
no. January, pp. 479–487, 2019, doi: [25]K. Schulze, M. Nagle, W. Spreer, B. Mahayothee, and J. Müller,
10.1016/j.compag.2019.01.032. “Development and assessment of different modeling approaches
[12]K. Hameed, D. Chai, and A. Rassau, “A comprehensive review of for size-mass estimation of mango fruits (Mangifera indica L.,
fruit and vegetable classification techniques,” Image Vis. cv. ’Nam Dokmai’),” Comput. Electron. Agric., vol. 114, pp.
Comput., vol. 80, 2018, doi: 10.1016/j.imavis.2018.09.016. 269– 276, 2015, doi: 10.1016/j.compag.2015.04.013.
[13] Tripathi, Mukesh Kumar, and Dhananjay D. Maktedar. [26]M. Fashi, L. Naderloo, and H. Javadikia, “The relationship
"Optimized deep learning model for mango grading: Hybridizing between the appearance of pomegranate fruit and color and size
lion plus firefly algorithm." IET Image Processing 15, no. 9 of arils based on image processing,” Postharvest Biol. Technol.,
(2021): 1940-1956. vol. 154, no. April, pp. 52–57, 2019, doi:
[14]X. Li et al., “Postharvest Biology and Technology SSC and pH for 10.1016/j.postharvbio.2019.04.017.
sweet assessment and maturity classi fi cation of harvested [27]Tripathi, Mukesh Kumar, Dhananjay Maktedar, D. N.
cherry fruit based on NIR hyperspectral imaging technology,” Vasundhara, CH VKNSN Moorthy, and Preeti Patil. "Residual
Postharvest Biol. Technol., vol. 143, no. May, pp. 112–118, Life Assessment (RLA) Analysis of Apple Disease Based on
2018, doi: 10.1016/j.postharvbio.2018.05.003. Multimodal Deep Learning Model." International Journal of
[15] Tripathi, M.K. and Maktedar, D.D., 2021. Detection of various Intelligent Systems and Applications in Engineering 11, no. 3
categories of fruits and vegetables through various descriptors (2023): 1042-1050.
using machine learning techniques. International Journal of [28]Y. Gurubelli, M. Ramanathan, and P. Ponnusamy, “Fractional
Computational Intelligence Studies, 10(1), pp.36-73. fuzzy 2DLDA approach for pomegranate fruit grade
[16]A. Bhargava and A. Bansal, “Fruits and vegetables quality classification,” Comput. Electron. Agric., vol. 162, no. March,
evaluation using computer vision: A review,” Journal of King pp. 95–105, 2019, doi: 10.1016/j.compag.2019.03.036.
Saud University - Computer and Information Sciences, 2018. [29]Nyalala et al., “Tomato volume and mass estimation using
[17]M. K. Tripathi and D. D. Maktedar, “A role of computer vision in computer vision and machine learning algorithms: Cherry
fruits and vegetables among various horticulture products of tomato model,” J. Food Eng., vol. 263, no. April, pp. 288–298,
agriculture fields: A survey,” Inf. Process. Agric., no. xxxx, 2019, doi: 10.1016/j.jfoodeng.2019.07.012.
2019, doi: 10.1016/j.inpa.2019.07.003. [30]V. O. Ayodele, O. M. Olowe, C. G. Afolabi, and I. A. Kehinde,
[18]Q. Kou, D. Cheng, L. Chen, and Y. Zhuang, “Principal curvatures “Identification, assessment of diseases and agronomic
based local binary pattern for rotation invariant texture parameters of Curcuma amada Roxb (Mango ginger),” Curr.
classification,” Optik (Stuttg)., vol. 193, p. 162999, Sep. 2019, Plant Biol., vol. 15, no. October, pp. 51–57, 2018, doi:
doi: 10.1016/j.ijleo.2019.162999. 10.1016/j.cpb.2018.10.001.
[19]Channapattana, Shylesha V., Srinidhi Campli, A. Madhusudhan, [31]M. Moalemiyan, A. Vikram, and A. C. Kushalappa, “Detection
Srihari Notla, Rachayya Arkerimath, and Mukesh Kumar and discrimination of two fungal diseases of mango (cv. Keitt)
Tripathi. "Energy analysis of DI-CI engine with nickel oxide fruits based on volatile metabolite profiles using GC/MS,”
nanoparticle added azadirachta indica biofuel at different static Postharvest Biol. Technol., vol. 45, no. 1, pp. 117–125, 2007,
injection timing based on exergy." Energy 267 (2023): 126622. doi: 10.1016/j.postharvbio.2006.08.020.
[20]G. P. Moreda, J. Ortiz-Cañavate, F. J. García-Ramos, and M. Ruiz- [32] Tripathi, M.K., Neelakantapp, M., Kaulage, A.N., Nabilal, K.V.,
Altisent, “Non-destructive technologies for fruit and vegetable Patil, S.N. and Bamane, K.D., 2023. Breast Cancer Image
size determination - A review,” J. Food Eng., vol. 92, no. 2, pp. Analysis and Classification Framework by Applying Machine
119–136, 2009, doi: 10.1016/j.jfoodeng.2008.11.004. Learning Techniques. International Journal of Intelligent
[21]F. S. A. Sa’ad, M. F. Ibrahim, A. Y. Md. Shakaff, A. Zakaria, and Systems and Applications in Engineering, 11(3), pp.930-941.
M. Z. Abdullah, “Shape and weight grading of mangoes using [33]H. Cen, Y. He, and M. Huang, “Measurement of soluble solids
contents and pH in orange juice using chemometrics and vis-

979
IJRITCC | October 2023, Available @ http://www.ijritcc.org
International Journal on Recent and Innovation Trends in Computing and Communication
ISSN: 2321-8169 Volume: 11 Issue: 10
Article Received: 11 August 2023 Revised: 19 September 2023 Accepted: 15 October 2023
___________________________________________________________________________________________________________________

NIRS,” J. Agric. Food Chem., vol. 54, no. 20, pp. 7437–7443,
2006, doi: 10.1021/jf061689f.
[34] Tripathi, M.K., Reddy, P.K., Neelakantappa, M., Andhare, C.V.,
Shivendra "Identification of mango variety using near infrared
spectroscopy”, Indonesian Journal of Electrical Engineering and
Computer Science this link is disabled, 2023, 31(3), pp. 1776–
1783.
[35]S. Wibowo, C. Buvé, M. Hendrickx, A. Van Loey, and T. Grauwet,
“Integrated science-based approach to study quality changes of
shelf-stable food products during storage: A proof of concept on
orange and mango juices,” Trends Food Sci. Technol., vol. 73,
no. November 2017, pp. 76–86, 2018, doi:
10.1016/j.tifs.2018.01.006.
[36]A. B. T. PETER BUTZ, CLAUDIA HOFMANN, “Recent
Developments in Noninvasive Techniques for Fresh Fruit and
Vegetable internal quality analysis,” J. Food Sci., vol. 70, no. 9,
pp. 131–141, 2005.
[37]E. J. N. Marques, S. T. De Freitas, M. F. Pimentel, and C. Pasquini,
“Rapid and non-destructive determination of quality parameters
in the ‘Tommy Atkins’ mango using a novel handheld near
infrared spectrometer,” Food Chem., vol. 197, pp. 1207–1214,
2016, doi: 10.1016/j.foodchem.2015.11.080.
[38] Shivendra, Chiranjeevi, K. and Tripathi, M.K., 2022. Detection of
Fruits Image Applying Decision Tree Classifier Techniques. In
Computational Intelligence and Data Analytics: Proceedings of
ICCIDA 2022 (pp. 127-139). Singapore: Springer Nature
Singapore.
[39] M. A. Hossain, M. M. Rana, Y. Kimura, and H. A. Roslan,
“Changes in biochemical characteristics and activities of
ripening associated enzymes in mango fruit during the storage at
different temperatures,” Biomed Res. Int., vol. 2014, 2014, doi:
10.1155/2014/232969.
[40]P. P. S. Gill, S. K. Jawandha, N. Kaur, and N. Singh, “Physico-
chemical changes during progressive ripening of mango
(Mangifera indica L.) cv. Dashehari under different temperature
regimes,” J. Food Sci. Technol., vol. 54, no. 7, pp. 1964–1970,
2017, doi: 10.1007/s13197-017-2632-6.
[41]D. Thinh, J. Uthaibutra, and A. Joomwong, “Effect of Storage
Temperatures on Ripening Behavior and Quality Change of
Vietnamese Mango Cv . Cat Hoa Loc,” Interntional J. Bio-
Technology, vol. 3, no. 3, pp. 19–30, 2013.
[42] Tripathi, M.K., 2023. Neutrosophic approach based intelligent
system for automatic mango detection. Multimedia Tools and
Applications, pp.1-23.s..
[43]G. Kansci, B. B. Koubala, and I. L. Mbome, “Biochemical and
physicochemical properties of four mango varieties and some
quality characteristics of their jams,” J. Food Process. Preserv.,
vol. 32, no. 4, pp. 644–655, 2008, doi: 10.1111/j.1745-
4549.2008.00204. x.

980
IJRITCC | October 2023, Available @ http://www.ijritcc.org