Automatic Object Detection and Dimensional Measurement of Object Using Image Processing
Automatic Object Detection and Dimensional Measurement of Object Using Image Processing
Volume 7, Issue 5, May – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Automatic Object Detection and Dimensional
Measurement of Object Using Image Processing
Anuradha Kambale1, Neha patil2, Neha Ranmale3, Shrutika Gonjare4
Professor1, Student2,3,4
Department of Electronics and Telecommunication
City- Palus, Country- India
Abstract:- In this days, automatic object detection and II. LITERATURE REVIEW
dimensions measurement is an important issue from so
many small scale industries. It automatically measured Nashwan Adman Othman, Mehmet Umut Salur, Mehnet
size of object with their high accuracy by structure of Karakose, Aydin all are form computer engineer They are
light is based on system of camera.Using the different belongs to Elazig and sanliurfa in Turkey country. They
types of algorithms as well as morphological operation design a project on embedded real-time object detection
and the combination of hardware and software. This is and measure the size of object. They use latest technology
the vital topic of computer vision problems this study for detecting the object and measure the size in real time for
present an enhanced technique for detecting object this purpose they use video of object. For object
dimension of the object. Quality of the product is measurement technique by using video they use various
becoming necessary and demands high precision with algorithm like OpenCV libraries include canny edge,
accuracy for each atom. The approach includes six steps detection, dilation, and erosion. And making a video and
namely (1) Training Image, (2) Image segmentation, (3) implementation of technique they design a system with the
feature explanation, (4) Testing of Image and the last is help of OpenCV, software library, Raspberry pi 3 and
(5) Classification. The grey scale image are firstly aligned Raspberry camera.
in the preprocessing step. In the implementation of the Cuong Vo-Le, Pham Van Muoi, Nguyen Hong Son,
purposed technique, we designed a system that used Nguyen Van San, Vu, Khac Duong and Nguyen ThiHuyen
MATLAB software, CCD/Web Camera. It nearly they all are from electronics and telecommunication
achieved 98% success in determination to size of object. department of Hanoi University Vietnam. Title of their
project is Automatic Method for Measurement Object Size
Keywords:- Object detection, dimensions measurement. Using 3D camera Main purpose of their project is measure
size of object with high accuracy using stereo camera
I. INTRODUCTION
system. This project is divided into four steplike
In this days, automatic object detection and dimensions preprocessing, object detection, key points extraction and
measurement is an important issue from so many small scale depth interpolation before size calculation. Then, the
industries. Because quality of the product is becoming project is detected by depth threshold and key points are
necessary and demands high precision with accuracy for each extracted by Shi-Tomasi corner detector combined with our
item. The small scale industry cannot invest the use amount proposed key points extraction algorithm.
of latest technology. They follow only traditional technique
III. OBJECTIVE
traditional instrument and they have limited capital for
expansion. This traditional/old technique are time A. To detect the object:
consuming. Project quality in industry depends on quality of project.
The system can be used and applied to an industrial quality
Manufacturing industry look and the some challenges as
control of the system. Our proposes system is helpful to
they have to deliver product at consistently high quality
detect the object by using different image processing
standards, navigate end-to-end supply chains and manage
techniques, so we first detect the object and apply various
strict to market deadline driven by demanding customers are
image processing techniques.
highly sensitive to time, cost, and quality. Study puts efforts
in effectiveness and accuracy for better quality through B. Study various defects:
Automatic image analysis is one of the basic approaches in The detection of products defect is essential in quality for
the field of industrial application, among other for inspection, manufacturing industry.
measurements or object detection. Nowdays, object detection
and dimension measurement of object is major issue for The main task is to detect surface irregularity means
many small scale industry. That’s why we trying to overcome check whether surface is smooth or rough also identities the
this challenge faced a small scale industry. And we develop a cracks on surface of project. In many industries manual
project is “Automatic Object Detection and Dimensional inspection of large metal surface is still in use even though is
Measurement Using Image Processing”. the precision of this method is restricted. Then the quality
requirement of the object have increased the particular with
respect to homogeneous flatness. It is used to detect the
surface and internal defects of projects. It is refers to detect
IJISRT22MAY753 www.ijisrt.com 1276
Volume 7, Issue 5, May – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
of the spot, pit, scratch, color difference and defects on the That ready to be shipped. He will be decide the product is
product surface. going to be tested. The newspaper (length, breath, width,
diameter, surface).
C. Identify problems faced by manufacturing industry: Inspection Box:Then inspection box collect the info all
To report third party and provide in time delivery having tested data of the product is transmit to the next block.
good quality of the product. We identify different quality Comparing data: Compare the tested data of the object
relevant problems such as meet the standard dimensions with standard data tested data of the object which is already
requirements as per product quality. Hence working is stored in data base. After compression this compared data
continuous and provide effective strategies. transmit to the signal box.
D. Reduce rejection of a product: Signal : The function of signal box is if tested data of the
The rejection level can be reduced by implementing the product is match with standard data of the product with
step by step approach in a manufacturing processing zero tolerance then tested product is OK otherwise if tested
providing good quality products which meet the standard data of the product dose not match with standard data of the
requirement of venture. This will result in reduction of
rejection and rework.
IV. METHODOLOGY
The proposed of model estimation the object detection
of individual objects from 2D images. This system is build
using object detection and dimensional measurement of
object detection. Initially, the methodology was developed
using only a single view of the object, however all the
necessary dimensions of the object cannot be computed using
only one view. So, the proposed model employs the use of
two views (any combination of front, top or side view)
depending upon the application. To avoid distorted
dimensions and obtain improved accuracy, image required
object are captured at a 900 viewing angle with respect to the
camera. At the present, the model is built for the following
object: Bricks, cube, paper.
The model consists of following tasks: object
recognition, dimension calculation. product with -1 to1 tolerance then tested product is not ok.
Result: Then last step is final result of that product is
transmit to Inspector .
Fig. A Cube
The process of this method automatically measuring Fig. B Brick
the object size based on image processing. The approach
includes six steps namely inspector, inspection box,
comparing data, signal object ok / not ok, result and last step
transmit feedback of the result to the Inspector.
Inspector: It is a first step of methodology. A inspector
monitors the quality aspects of incoming product.
IJISRT22MAY753 www.ijisrt.com 1277
Volume 7, Issue 5, May – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
V. FLOW CHART
Fig. 1: Flow Chart Fig. 2: Block Diagram
Training image- A bunch of image for which the required VI. CONCLUSION
outcome is know for which the required outcome is known.
You input them to a program that analysers their features Now days object detection and dimension measurement
through a classification routine that determines the of object is major issue for many small scale industry.
appropriate data to use on the features in order to best
achieve the required result. So, that trying to overcome this challenges faced a small
Image segmentation- It is a method in which a digital scale industry.
image is broken down into various subgroups. - image
segmentation is a commonly used technique in digital So, we are developing the Automatic Object Detection
image processing Image segmentation is the process by and Dimensional Measurement using image processing.
which a digital image is Partitional into various subgroups Manually operated instrument for dimension inspection
(of pixel) called image objects, which can reduce the which is caliber and time consuming which affect to overall
complexity of the image, and thus analysing the image industrial manufacturing process. Hence it is prepared of
become simpler. developing “Automatic Object Detection and Dimensional
Feature explanation- In computer vision and image Measurement Using Image Processing” which will be more
processing, a feature is a piece of information about the precise for batter quality produces.
content of an image.In computer vision and image REFERENCES
processing, a feature is a piece of information about
thecontent of an image; typically about whether a certain [1.]M. Leo, A. Natale, M. Del-Coco, P. Carcagni, and C.
region of the image has certain properties. Features may be Distante, “Robust estimation of object dimensions and
specific structures in the image such as points, edges or external defect detection with a low-cost sensor,” Journal
objects. of Nondestructive Evaluation, vol. 36, no. 1, p. 17, 2017.
Testing of image- Compare the two different images.A [2.]Cuong Vo-Le, Pham Van Muoi, Nguyen Hong Son, 2020
standard test image is a digital image file used across IEEE Eighth International Conference on
different institutions to test image processing and image Communications and Electronics (ICCE) | 978-1-7281-
processing and image compression algorithms. By using 5471-8/20 2021 IEEE | DOI:
the same standard test images, different labs are able to 10.1109/ICCE48956.2021.9352115
compare results, both visually and quantitatively. [3.]Nashwan Adnan, Mehmet Umut 978-1-5386-6878-8/18
Feature extraction- Feature extraction is a part of the ©2018 IEEE.
dimensionality reduction process.
Classification- Image classification is a complex procedure
which relies on different components. Digital image
classification uses the spectral information represented by
the digital numbers in one or more spectral bands, and
attempts to classify each individual pixel based on this
spectral termed spectral pattern recognitions. Two general
methods of classification are ‘supervised’ and
‘unsupervised’.
