Application of image processing in agriculture

6. Introduction:

The Agriculture is one of the largest economic sectors and it plays the
major role in economic development of our country. In our country the ever-increasing
population, losses involved in processing and the increasing demand of fruits of high
quality with good appearance, there is a need for the development of accurate, fast and
focused quality determination of food and agricultural products like fruits and
vegetables. Handling process of agricultural produce is performed in several steps like
first cutting of fruits or vegetables from the farm, washing, sorting, grading, packing,
transporting and finally storage. Amongst all these steps sorting and grading are major
processing tasks associated for preserving the quality of fresh-market stuff. Sorting of
agricultural products is done based on appearance of fruits. Whereas grading is done
based on the overall quality features of a fruits by considering a number of attributes
like shape, size, color etc.
Fruit market is getting highly selective, requiring their suppliers to
distribute the fruits of high standards of quality and presentation as well. So there is a
increasing need to supply quality fruits within a short period of time has given rise to
the development of automated Grading of fruits to improve the quality.

A computer vision system is a cost effective system and gives consistent

performance, a superior speed and accurate sorting and grading of fruits. Computer
vision based sorting and grading had undergone substantial growth in the field of
agricultural sector in the developed and developing countries because of availability of
the infrastructures. Computer vision is the construction of explicit and meaningful
descriptions of physical objects from images.

7. Literature Review :
1. M. R. Satpute and S. M. Jagdale, "Automatic fruit quality inspection system," 2016
(ICICT), 2016:[1]
In this paper, An automatic fruit quality inspection system for sorting and grading of
tomato fruit and defected tomato detection has been presented. The main aim of this
system is to replace the manual inspection system. This helps in speed up the process
improve accuracy and efficiency and reduce time. This system collect image from camera
which is placed on conveyor belt. Then image processing is done to get required features
of fruits such as texture, color and size. Defected fruit is detected based on blob detection,
color detection is done based on thresholding and size detection is based on binary image
of tomato. Sorting is done based on color and grading is done based on size.

In this approach, 30 of 35 red tomatoes, 33 of 35 green tomatoes, 30 of 35 yellow

tomatoes and27 of 30 defected tomatoes are detected correctly. The variation in speed of
conveyor and light, camera resolution affects the system. The accuracy of green color
detection is 94.28% which more than red and yellow. The accuracy of defect fruit
detection is up to 90%.This system is much closer to manual expert judge.

In the Further design, this system can be modified by increasing size of conveyor belt so
that it is possible to perform quality inspection of large fruit than tomato, and increase
accuracy of the system so that it can differentiate between artificial , hybrid color from
original fruit color.

2. Ruchita R. Mhaski P.B. Chopade M.P. Dale, “Determination of Ripeness and

Grading of Tomato using Image Analysis on Raspberry Pi”, International Conference
on Communication, Control and Intelligent Systems (CCIS), 2015 (IEEE)
Presently, human experts grade the agriculture goods based on its vision based features
that cause the inaccuracy, inconsistency and inefficiency on defining the quality of
agriculture goods. In this paper, we are inspecting the quality of tomato based on shape,
size and degree of ripeness. An edge detection algorithm is used to estimate the shape and
 size of tomato and color detecting algorithm is used for the ripeness determination. All
these algorithms are implemented on Raspberry Pi development board which will
become independent and cost effective system. Our system includes Raspberry Pi
development board, conveyor belt, motors, Pi camera. All the interfacing of the above
devices will be carried out and will make a cost effective embedded system for the
determination of shape, size and degree of ripeness of tomato. Same system can be
utilized for other fruits and vegetables also.

This paper proposed and appraised the frame for the determination of the ripeness of
tomato, estimation of size and shape and detection of defects of tomato. This work
presents the novel incorporated technique for grading tomatoes based on their ripeness.
The proposed approach used the concept of color detection algorithm for classifying
tomato in three classes and K-mean clustering algorithm for detecting the defects in
tomato and classifying them in defected class. The result analysis is able to accurately
determine the ripeness of tomato. The modularity and distributed nature of approach
makes the system easy to be upgraded in the future in order to increase quality standard.

3. Megha P. Arakeria, Lakshmanab, “Computer Vision Based Fruit Grading System for
Quality Evaluation of Tomato in Agriculture industry”, 7th International Conference
on Communication, Computing and Virtualization 2016.

The automated grading system not only speeds up the time of processing, but also
minimizes error. There is a great demand for tomatoes in both local and foreign markets.
The tomato fruit is very delicate and hence careful handling of this fruit is required
during grading. Thus, this paper proposes an automatic and effective tomato fruit grading
system based on computer vision techniques. The proposed quality evaluation method
consists of two phases: development of hardware and software. The hardware is
developed to capture the image of the tomato and move the fruit to the appropriate bins
 without manual intervention. The software is developed using image processing
techniques to analyze the fruit for defects and ripeness.

Experiments were carried out on several images of the tomato fruit. It was observed that
the proposed method was successful with 96.47% accuracy in evaluating the quality of
the tomato.

4. Vilas D. Sadegaonkar, Kiran H. Wagh, " Automatic Sorting Using Computer Vision
& Image Processing For Improving Apple Quality," International Journal of
Innovative Research and Development: 2015

In agricultural sector the efficiency and the accurate grading process is very essential to
increase the productivity of produce. Everyday high quality fruits are exported to other
countries and generate a good income. That is why the grading process of the fruit is
important to improve the quality of fruits. However, fruit grading by humans in
agricultural industry is not sufficient, requires large number of labors and causes human
errors. Automatic grading system not only speeds up the process but also gives accurate
results. Therefore, there is a need for an efficient fruits grading or classification methods
to be developed. Fruit’s color, size, weight, component texture, ripeness are important
features for accurate classification and sorting of fruits such as oranges, apples, mangoes
etc. Objective of this paper is to emphasize on recent work reported on an automatic fruit
quality detection system. This paper presents the image processing techniques for feature
extraction and classification for fruit quality measurement system.

8. Concluding Remark :

• The most of the existing system are based on the color shape and size.
• The system may get failed for hybrid color of fruit
• The most of the vision based systems were failed due to intensity variations
• In the proposed system, we planned to develop the machine learning based system
 • The solution for portability of the hardware can be solved by the using tiny
Raspberry Pi.

9. Problem Definition:
Currently, the fruits according to their ripeness and quality are sorted manually
by human. In this process due to human error the fruits can be sorted wrongly which
leads to wastage of fruit.

10. Objectives of the present work:

1. Literature survey of different approach to design image processing based fruit sorter.
2. Creation of database
3. To pre-process the video data sequence to make it suitable for further processing
Using noise removal technique like Median, Gaussian filtering etc.
4. To implement feature extraction algorithms
5. To analyze the effectiveness and usefulness of classification techniques
6. Implementation of the hardware section
7. To validate the performance of the system and compared with exiting system

11. Proposed System:

Fig. 1 and Fig 2 give an idea of how the system design is proposed. The
conveyor belt which will run on DC motors will bring the fruit in front of the camera.
Image processing will do on the image of the fruit. According to the features of the
image, it will be categorized into 3 types – raw, half ripe and ripe. From the resulted
category of the fruit it will be given to the sorting assembly which will sort the fruits.

The following steps briefly define the working of the device:

1. The fruit is transported to the front of the camera.
2. Image of the fruit is taken.
3. Image is stored and then is processed which will be categorized.
 4. According to the category of the fruit, the fruit will be sorted by the sorting
assembly.

Block Diagram:

Camera Raspberry Pi Alert

Fig.1 Block diagram

 Fruit image training Fruit image testing
database database

Preprocessing Preprocessing

Image Image
segmentation segmentation

Feature Feature
extraction extraction

Trained Database Classifier

STOP Grade II
Grade I Grade III

Fig.2 Image processing section

The information of the modules and component to be use in the system is explain below

• Raspberry Pi:

The raspberry Pi is a controller board based on the Broadcom BCM2837 chip

which is used in many mobile phones. The BCM2837 contains an 4× ARM Cortex-A53,
 processor running at 1.2GHz, 1GB LPDDR2 (900 MHz) of RAM and a GPU named
Broadcom Video Core IV. The raspberry pi2 board runs on 5V 2A DC supply and
consists of 4 USB ports, 3.5mm analogue audio-video jack, Ethernet, Camera Serial
Interface (CSI), and Display Serial Interface (DSI).

• Camera

The custom made camera RaspiCam for Raspberry Pi or any webcam.

• Motors
DC motors will be used for the movement of the conveyor belt with 12v DC
input.

• Mechanical assembly
Mechanical assembly will contain a conveyor belt for transporting the fruit to the
sorting assembly.

12. Scope of work :

• In the image processing based project, the use of Raspberry Pi becomes

advantageous because it provides a great feature.
• Interfacing of Camera and motors with Raspberry Pi.
• To design the mechanical fruit sorting system.
• Evaluate the performance of the proposed system.

13. Methodology [Action Plan] :

The proposed work will be carried out with following steps:

1. Phase 1: To carry out the detailed literature survey regarding the image processing
based fruit sorter systems.
 2. Phase 2: The actual arrangement of system hardware and software.

3. Phase 3: Testing of system to get the results.

4. Phase 4: Work is concluded with suggestion for future work.

14. Facilities available:

Facilities are available at parent institute,

• Central library well equipped with reference books, latest electronics magazines.

• Digital library of having access of IEEE and science software.

17. Expected date of completion : April, 2018

18. Approximate Expenditure : 10000 Rs.

19. Reference:

[1] M. R. Satpute and S. M. Jagdale, "Automatic fruit quality inspection system," 2016
International Conference on Inventive Computation Technologies (ICICT), Coimbatore,
2016, pp. 1-4.
[2] R. R. Mhaski, P. B. Chopade and M. P. Dale, "Determination of ripeness and grading of
tomato using image analysis on Raspberry Pi," 2015 Communication, Control and
Intelligent Systems (CCIS), Mathura, 2015, pp. 214-220.
[3] Megha. P. Arakeri, Lakshmana, “Computer Vision Based Fruit Grading System for
Quality Evaluation of Tomato in Agriculture industry”, 7th International Conference on
Communication, Computing and Virtualization 2016, pp. 426 – 433.
[4] Dr. Vilas D. Sadegaonkar, Kiran H. Wagh, “Automatic Sorting Using Computer Vision
& Image Processing For Improving Apple Quality”, International Journal of Innovative
Research and Development, Vol. 4. Issue 1, Jan 2015, pp.11-14