ABSTRACT

Object dimension detection using OpenCV involves leveraging image processing techniques
to measure the physical dimensions of objects in images or video. The process typically starts
with acquiring a well-lit, high-resolution image containing the object and a reference object of
known size. OpenCV functions like edge detection and contour finding are used to isolate the
object. The pixel-to-real-world unit conversion is calculated using the reference object. By
determining the object's boundaries in pixels and applying the conversion factor, its
dimensions (length, width, and height) can be accurately estimated. This method is widely
used in automation, manufacturing, and quality control for non-contact measurement.Object
dimension detection using OpenCV involves identifying and measuring the size of objects in
an image or video. By leveraging edge detection, contour detection, and perspective
transformations, OpenCV can extract object dimensions in real-world.

ⅳ
 TABLE OF CONTENTS

ABSTRACT ⅳ
TABLE OF CONTENTS ⅴ
LIST OF FIGURES ⅵ
1 INTRODUCTION 7-10
1.1 PROBLEM STATEMENT 7
1.2 SOLUTION PROVIDED 8
1.3 OBJECTIVE 9
1.4 FUTURE ASPECT OF THE PROJECT 9
2 LITERATURE SURVEY 11-13
3 REQUIREMENTS AND SPECIFICATIONS 14-17
3.1 PROGRAMING LANGUAGE AND 14
DEVELOPMENT ENVIRONMENT
3.1.1 PROGRAMMING LANGUAGE 14
3.1.2 DEVELOPMENT ENVIRONMENT 15
3.2 SYSTEM REQUIEMENTS 15
3.2.1 SOFTWARE REQUIEMENTS 16
3.2.2 HARDWARE REQUIEMENTS 16
4 METHODOLOGIES 18-21
4.2 EXISTING SYSTEM 18
4.1 PROPOSED SYSTEM 20
5 SOFTWARE REQUIREMENT SPECIFICATIONS 22-23
5.1 FUNCTIONAL REQUIREMENT 22
5.2 NON- FUNCTIONAL REQUIREMENT 23
6 TESTING 24-26
6.1 SOFTWARE TESTING 24
6.2 HARDWARE TESTING 25
7 INTERPRETATION OF RESULTS 27
CONCLUSION 28
REFERENCE 29

ⅴ
 LIST OF FIGURES

7.1 REAL TIME DIMENSION DETECTION USING 28

CAMERA
7.2 REAL TIME DIMENSION DETECTION USING 28
CAMERA

ⅵ
 Object Dimension Detection Using Open cv

CHAPTER 1

INTRODUCTION

Object dimension detection involves determining the physical dimensions (e.g., length,
width, and sometimes height) of objects from images or video frames. This technique
is essential in fields such as manufacturing, logistics, quality control, and augmented
reality, where precise measurements are critical.

OpenCV, a widely-used open-source computer vision library, offers a robust set of

tools to implement object dimension detection. By leveraging image processing
techniques like edge detection, contour analysis, and perspective transformations,
OpenCV can extract the outlines of objects and calculate their dimensions in pixel
units.

To convert these pixel measurements into real-world units, a reference object with
known dimensions is often included in the image for calibration. This enables the
scaling factor to be determined, ensuring accurate and consistent results. The method
provides a non-contact, cost-effective alternative to traditional physical measuring
tools, making it highly versatile for a variety of applications.

In this process, tasks such as image preprocessing, feature detection, and geometric
transformations are crucial to achieve reliable results. Combined with its adaptability
to different environments, OpenCV's capabilities make it an ideal choice for object
dimension detection

1.1 PROBLEM STATEMENT

“Develop a system to accurately measure and detect the dimensions (length, width,
height) of physical objects using computer vision techniques. This solution should
enable automated dimension detection for applications in logistics, manufacturing,
and quality control.”

The goal of this project is to develop a real-time system that can measure the dimensions
(width and height) of objects captured by a webcam and display these dimensions in

CSE Dept, JCET HUBBALLI Page | 7

 Object Dimension Detection Using Open cv

centimeters. The system utilizes OpenCV to process live video feed from the webcam,
detect objects within the frames using edge detection and contour finding, and then
calculate the real-world dimensions of those objects. By using a reference object of
known size, the system calculates the Pixels per Centimeter (PPC) ratio and converts the
detected object’s dimensions from pixels to centimeters. The system will display
bounding boxes around the objects in the video feed, showing their calculated dimensions
in real time. This solution has applications in fields such as inventory management,
robotics, and quality control, where accurate object measurements are essential. However,
the accuracy of measurements depends on correct calibration with the reference object
and the conditions under which the system operates, such as lighting and object visibility.
The system is designed to process frames in real-time, ensuring efficient and smooth
performance while handling various object shapes and sizes in the video feed.

1.2 SOLUTION PROVIDED

This project provides real-world solutions across multiple industries by automating the
measurement of object dimensions in real time. In logistics and supply chain
management, it optimizes shipping and storage by accurately measuring packages,
calculating volumetric weight, and enhancing sorting and arrangement efficiency. E-
commerce businesses benefit from precise packaging, reduced returns, and faster
processing of orders and returns. In manufacturing, the system ensures quality control
by verifying product dimensions and detecting defects, streamlining production lines.
Retailers and warehouses can use it for inventory management, recording item
dimensions, optimizing storage, and conducting swift stock audits. Construction and
engineering teams can measure materials and components on-site, ensuring accurate

dimensions for assembly or installation. In healthcare, it facilitates precise sizing of

medical equipment or tools, and in food production, it ensures compliance with
packaging standards and uniformity. Additionally, the system supports education and
research, offering a practical tool for teaching image processing and enabling
experiments requiring non-contact measurements. Overall, the project enhances
efficiency, accuracy, and scalability, delivering significant cost savings and reducing
manual labor across various domains.

CSE Dept, JCET HUBBALLI Page | 8

 Object Dimension Detection Using Open cv

1.3 OBJECTIVE
The primary objective of this project is to develop a real-time, automated system for
detecting and measuring the dimensions of objects using computer vision techniques. By
leveraging a standard webcam and image processing algorithms, the project aims to
provide accurate and efficient measurement solutions to replace traditional manual
methods.

The specific goals include:

1. Accuracy: Ensuring precise measurement of object dimensions by calibrating

pixels per centimeter (PPC) based on a reference object.
2. Automation: Eliminating the need for manual tools or human intervention by
automating the dimension detection process.
3. Efficiency: Enabling real-time processing to measure multiple objects quickly in
dynamic environments.
4. Scalability: Designing a system that can be adapted for various object sizes and
applications across industries.
5. Cost-Effectiveness: Providing a low-cost alternative to expensive
commercial dimensioning systems.
6. User-Friendly Design: Developing a system that is simple to operate and can
integrate seamlessly into existing workflows.

1.4 FUTURE ASPECTS OF THE PROJECT

1. Enhanced Accuracy with Improved Edge Detection

Future advancements could focus on refining the edge detection algorithm to provide
more accurate object boundaries. Techniques like adaptive thresholding or machine
learning- based edge detection could be incorporated to enhance the precision of
dimension measurements in 2D.

2. Real-Time Calibration for Variable Object Distances

Implementing automatic real-time calibration will allow the system to account for objects
at varying distances from the camera, ensuring consistent and accurate measurements in
2D, regardless of the object’s location within the frame.

CSE Dept, JCET HUBBALLI

Page | 9
 Object Dimension Detection Using Open cv

3. Mobile and User-Friendly Interface

Developing a mobile application or an intuitive graphical user interface (GUI) would

make the system more accessible. Users could easily capture and measure dimensions on-
the-go, increasing the versatility and usability of the project for real-world applications.

4. Optimized Edge Detection and Noise Reduction

Future improvements in edge detection algorithms and noise reduction techniques (like
advanced Gaussian blur or edge-preserving filters) could further refine the system’s
ability to detect clear object boundaries in various lighting and environmental conditions.

CSE Dept, JCET HUBBALLI

Page | 10
 Object Dimension Detection Using Open cv

CHAPTER 2
LITERATURE SURVEY

The real-time object dimension detection project uses computer vision techniques to
detect and measure the dimensions (e.g., length and width) of objects in real time. The
following literature survey provides an overview of relevant research, technologies, and
methodologies in the field of computer vision, object dimensioning, and related areas,
highlighting advancements and gaps that this project aims to address.

Object Detection Overview:

 Object detection has evolved significantly with the introduction of machine

learning and deep learning techniques. Traditional methods of object detection
relied on feature extraction (e.g., SIFT, HOG) and classifiers like SVMs.
However, recent advancements in convolutional neural networks (CNNs),
particularly YOLO (You Only Look Once) and SSD (Single Shot Multibox
Detector), have drastically improved object detection capabilities, making it
feasible to detect objects in real-time applications.
 Object detection techniques are used extensively for dimension measurement. In
these systems, once an object is detected, its dimensions are estimated by
analyzing its boundary using bounding boxes, contours, or more advanced
techniques such as region proposals and polygon fitting.

Real-Time Object Detection:

 Real-time object detection is essential for applications like robotics, surveillance,

and industrial automation. Research such as the work by Redmon et al. on YOLO
demonstrates a significant leap in real-time object detection by allowing accurate
detection of objects in images at high speeds.

CSE Dept, JCET HUBBALLI

Page | 11
 Object Dimension Detection Using Open cv

Edge Detection:

 Edge detection is one of the fundamental steps in object measurement systems.

Canny edge detection, Sobel filters, and Laplacian of Gaussian are widely used
algorithms for detecting object edges. These edges form the boundaries of an
object and are critical for determining its shape and size.
 Canny edge detection, in particular, is used to detect a wide range of edges in
noisy environments. The multi-threshold approach of the Canny detector helps
to identify strong and weak edges, improving detection accuracy.

Contour Detection:

 Contour detection algorithms help in finding the outline of an object after edge
detection. OpenCV's findContours() is commonly used for this purpose.
Once the contours are detected, a bounding box is drawn around the object, and
its height and width are used as the dimensions of the object.
 Studies have shown that bounding boxes are effective for rectangular or square
objects, but more advanced techniques are required for irregular shapes. For
instance, fitting ellipses or polygons can offer more accurate measurements for
non-rectangular objects.

Camera Calibration:

 Camera calibration is essential in converting pixel measurements to real-world

units (like centimeters or inches). Accurate calibration allows the system to scale
pixel-based measurements into real-world dimensions. Research has focused on
techniques for calibrating cameras using reference objects with known
dimensions.
 One widely used calibration method is Zhang’s method, which involves using a
checkerboard pattern to calculate camera intrinsic and extrinsic parameters,
ensuring that measurements are accurate across different angles and distances.

CSE Dept, JCET HUBBALLI

Page | 12
 Object Dimension Detection Using Open cv

Reference Object-Based Calibration:

 In many object dimensioning systems, a reference object with known dimensions

(like a piece of paper, business card, or a specific reference marker) is used to
calibrate the camera. By calculating the pixels per centimeter (PPC) or pixels
per inch (PPI) ratio using this reference object, the system can then infer the
dimensions of other objects in the scene.
 Studies have shown that the accuracy of object measurement heavily relies on the
precision of the calibration process. Any distortion in the reference object (due to
camera angle, lens distortion, etc.) can affect the overall measurement accuracy.

AI for Object Recognition and Dimensioning:

 Recent literature suggests integrating artificial intelligence (AI) and machine

learning techniques to improve object recognition and dimension prediction.
Deep learning models, particularly CNNs, can be trained on large datasets to
predict object dimensions based on learned features.

CSE Dept, JCET HUBBALLI

Page | 13
 Object Dimension Detection Using Opencv

CHAPTER 3

REQUIREMENTS AND SPECIFICATIONS

The Object Dimension Detection project requires essential hardware components, including a
webcam or camera for capturing real-time images and a reference object for calibration to
ensure accurate measurements. The software setup involves Python as the programming
language and OpenCV for image processing and computer vision tasks, supported by a
development environment like Visual Studio Code. The system also needs adequate
computational resources to process frames efficiently, with a recommendation for a
system equipped with a modern processor and sufficient RAM. Specifications include the
ability to detect object dimensions with minimal error, a minimum contour area threshold
for filtering noise, and real-time frame processing to display results dynamically. The
project aims for user-friendly implementation, enabling scalability for various
applications across industries like manufacturing, retail, and logistics. The requirements
include OpenCV, a camera or imaging device, proper lighting, and a reference object of
known size. The system must accurately detect, isolate, and measure object dimensions in
images.

3.1 PROGRAMMING LANGUAGE AND DEVELOPMENT

ENVIRONMENT

The project utilizes a robust combination of programming language and development tools to
deliver an efficient and effective system. The details are outlined below:

3.1.1 Programming Language:

Python was selected as the primary programming language for this project because
of its simplicity and versatility, which make it ideal for a wide range of applications. Its
straightforward syntax and rich ecosystem of libraries, such as OpenCV for image
processing ensured that the project could be developed efficiently. Additionally, Python's
cross-platform compatibility allowed the solution to work seamlessly across different
operating systems.

CSE Dept, JCET HUBBALLI Page | 14

 Object Dimension Detection Using Opencv

The development environment included Python's Default IDE (IDLE) and Visual Studio
Code (VS Code). IDLE was used for quick prototyping and small-scale testing, while VS
Code was employed for the more structured aspects of the project, offering features like
debugging tools, code completion, and Git integration. This combination of Python and
development environments facilitated an effective and efficient workflow throughout the
project.

3.1.2 Development Environment:

The development of the project was carried out using Python's Default IDE (IDLE) and
Visual Studio Code (VS Code). IDLE provided a simple and accessible platform for
quick prototyping and testing, with features like an interactive Python shell and built-in
syntax highlighting that enhanced the coding experience. Its lightweight nature made it
ideal for debugging and small-scale testing tasks.

Visual Studio Code was utilized for more comprehensive development needs. It
offered advanced capabilities such as integrated debugging, intelligent code completion,
and Python-specific extensions like PyLint. The integrated terminal facilitated seamless
script execution, while Git integration supported effective version control and
collaboration. The flexibility and scalability of VS Code made it a powerful tool for
managing and executing the project's larger components. Together, these environments
ensured an efficient and productive workflow for the project.

This development setup ensures the project's technical reliability and scalability while
maintaining ease of use for end-users.

3.2 SYSTEM REQUIREMENT

To develop a system for object dimension detection using OpenCV, you'll need to
consider the software and hardware requirements that meet the computational and
application demands. Below is a detailed guide:

CSE Dept, JCET HUBBALLI Page | 15

 Object Dimension Detection Using Opencv

3.2.1 Software Requirements:

 Operating System

 Windows, Linux, or macOS

 A 64-bit OS is recommended for better performance and compatibility.

 Programming Environment

 Python or C++: Python is easier for beginners and has better library support,
while C++ offers higher performance.

 Libraries

 OpenCV: Core library for computer vision tasks.

 Numpy: For mathematical operations and matrix handling.

 IDE or Text Editor

 Visual Studio Code, PyCharm, Jupyter Notebook (for Python development), or

CLion (for C++ development).

3.2.2 Hardware Requirements:

1. Processor

 A multi-core processor with at least 2 GHz clock speed (e.g., Intel i5 or AMD
Ryzen 5) is recommended.
 For faster processing, an Intel i7/i9, AMD Ryzen 7/9, or equivalent.

2. Memory (RAM)

 Minimum: 4 GB
 Recommended: 8 GB or higher for handling real-time processing or high-
resolution images.

CSE Dept, JCET HUBBALLI Page | 16

 Object Dimension Detection Using Opencv

3. GPU (Optional, but recommended for large-scale systems)

 If the application involves deep learning or real-time processing, a GPU such as:
o NVIDIA GeForce GTX/RTX series or equivalent with CUDA support.
o AMD GPUs with OpenCL support.

4. Camera

 A high-resolution camera (e.g., 1080p or 4K) for precise measurements.

 Frame rate: Minimum 30 FPS for real-time processing.
 Depth cameras (e.g., Intel RealSense or Kinect) can be useful for 3D
measurements.

5. Storage

 SSD with at least 256 GB for faster file access and system responsiveness.

6. Peripheral Devices

 Measuring tools for calibration (e.g., a checkerboard or Aruco markers).

 Tripod or stabilizer for camera stability.

CSE Dept, JCET HUBBALLI Page | 17

 Object Dimension Detection Using Opencv

CHAPTER 4
METHODOLOGIES

4.1 EXISTING SYSTEM

1. Camera Initialization

 Purpose: Open a connection to the webcam for capturing real-time video.

 Implementation:
o cv2.VideoCapture(1) initializes the second connected camera
(index 1).
o Checks if the webcam is successfully opened (cap.isOpened()).

2. Calibration for Real-World Measurements

 Purpose: Convert pixel dimensions to real-world units (centimeters).

 Implementation:
o A reference object with known dimensions is used (21 cm × 29.8 cm).
o The reference width in pixels is determined, and a conversion factor
(Pixels Per Centimeter, or PPC) is calculated.

3. Frame Preprocessing

 Purpose: Prepare each video frame for contour detection.

 Steps:
o Convert the frame to grayscale (cv2.cvtColor) to simplify the analysis
by focusing on intensity.
o Apply Gaussian blur (cv2.GaussianBlur) to reduce noise and smooth
the image.
o Perform edge detection (cv2.Canny) to identify object boundaries clearly.

4. Contour Detection

 Purpose: Identify the shapes of objects in the image based on their edges.
 Implementation:
o Find external contours using cv2.findContours.

CSE Dept, JCET HUBBALLI Page | 18

 Object Dimension Detection Using Opencv

5. Filtering Contours

 Purpose: Remove noise and irrelevant small objects.

 Implementation:
o Check the area of each contour using cv2.contourArea.
o Only process contours that are larger than a predefined
threshold (MIN_CONTOUR_AREA).

6. Bounding Box Calculation

 Purpose: Determine the rectangle enclosing each object of interest.

 Implementation:

o Compute the bounding rectangle for each valid contour using

cv2.boundingRect.
o Extract the rectangle's width (w) and height (h) in pixels.

7. Real-World Dimension Estimation

 Purpose: Convert pixel dimensions to real-world measurements.

 Implementation:
o Use the previously calculated PPC to convert the width (w) and height (h)
from pixels to centimeters.

8. Overlay Annotations

 Purpose: Provide visual feedback on the frame.

 Implementation:
o Draw bounding boxes around objects using cv2.rectangle.
o Display the calculated dimensions on the frame using cv2.putText.

9. Real-Time Processing Loop

 Purpose: Continuously process video frames until a termination condition is met.
 Implementation:

o The loop captures frames, processes them, and displays the result in a
window.

CSE Dept, JCET HUBBALLI Page | 19

 Object Dimension Detection Using Opencv

10. Resource Cleanup

 Purpose: Release resources and close all display windows when the program
ends.
 Implementation:
o Release the webcam using cap.release().
o Close all OpenCV windows using cv2.destroyAllWindows().

4.2 PROPOSED SYSTEM

The proposed system focuses on overcoming the challenges of the existing methodology,
enhancing the robustness, accuracy, and usability of the object dimension detection
system.

1. Camera Calibration for Precision

 Current State: Uses a fixed PPC value based on a reference object.

 Proposed Improvement:

o Implement full camera calibration using a checkerboard

pattern (cv2.calibrateCamera()).
o Correct lens distortion with cv2.undistort().
o Dynamically calculate PPC based on camera calibration and distance
from the object.

2. Perspective Correction

 Current State: Assumes the object is aligned with the camera plane.
 Proposed Improvement:
 Use a perspective transformation (cv2.getPerspectiveTransform()).
o Add a method to align objects to a frontal plane, ensuring accurate
dimension measurement regardless of camera angle.

3. Adaptive Thresholding

 Current State: Uses fixed thresholds for Canny edge detection.

CSE Dept, JCET HUBBALLI Page | 20

 Object Dimension Detection Using Opencv

 Proposed Improvement:
o Replace fixed thresholds with adaptive
methods (cv2.adaptiveThreshold() or auto-
calibrated Canny).
o Dynamically adjust thresholds based on image intensity and contrast levels.

4. Dynamic Contour Filtering

 Current State: Filters contours using a fixed MIN_CONTOUR_AREA.

 Proposed Improvement:
o Introduce a dynamic area threshold based on the frame's resolution and
scene complexity.
o Allow user-defined parameters to adapt to different use cases.

CSE Dept, JCET HUBBALLI Page | 21

 Object Dimension Detection Using Opencv

CHAPTER 5
SOFTWARE REQUIREMENTS SPEACIFICATIONS

5.1 FUNCTIONAL REQUIREMENTS

1. Webcam Access:
- The system must access and capture video from a webcam connected to the system.
- The program must be able to open the webcam and stream frames in real-time.

2. Frame Capture and Processing:

- The system must continuously capture frames from the webcam feed.
- The frames must be processed (grayscale conversion, Gaussian blur, edge detection).

3. Edge Detection:
- The system should apply an edge detection technique (e.g., Canny edge detection) to
the captured frames to identify the boundaries of objects.

4. Contour Detection:
- The system should identify and extract contours from the edge-detected frame to
detect objects.

5. Contour Filtering:
- The system should filter out small contours (based on a predefined area threshold) to
avoid detecting irrelevant or noise objects.

6. Bounding Box Drawing:

- For each relevant contour (object), a bounding box should be drawn around it,
highlighting the detected object in the frame.

7. Dimension Calculation:
- For each detected object, the system must calculate its dimensions (width and height)
in centimeters, based on a reference object with known dimensions.

CSE Dept, JCET HUBBALLI Page | 22

 Object Dimension Detection Using Opencv

8. Display Object Dimensions:

- The system must overlay the calculated object dimensions (in centimeters) on the
frame, preferably near the object’s bounding box.

9. User Interaction:
- The system must allow the user to exit the video stream by pressing the 's' key.

10. Real-Time Display:

- The system should display the processed frames with bounding boxes and dimensions
in real-time on the screen.

5.2 NON-FUNCTIONAL REQUIREMENTS

1. Performance:
- The system should process frames in real-time with minimal lag.
- The detection and dimension calculation should happen quickly enough for a
smooth user experience.
2. Accuracy:
- The accuracy of dimension calculation should be high, with minimal error based on
the reference object dimensions.
- The system should be able to detect objects accurately and reliably, with minimal
false positives or negatives.
3. Scalability:
- The system should be able to detect objects of varying sizes within the camera’s
field of view, depending on the distance of the object from the camera.
4. Usability:
- The user interface (display) should be simple and easy to understand, showing the
dimensions clearly and intuitively.
- The system should be able to handle typical object types without requiring
specialized training or complex configurations.

CSE Dept, JCET HUBBALLI Page | 23

 Object Dimension Detection Using Opencv

CHAPTER 6
TESTING THE CODE

To perform software and hardware testing for your object dimension detection system, you
need to verify its performance, accuracy, and stability under various conditions. Below is
a guide on how to systematically test your code:

6.1. SOFTWARE TESTING

a. Functional Testing

 Objective: Ensure that all functionalities (e.g., contour detection, measurement

calculations) work as expected.
 Steps:
1. Test with objects of known dimensions (e.g., a box or a sheet of paper)
and compare calculated dimensions to the actual dimensions.
2. Validate the PPC calibration logic.
3. Check how the program handles small objects (less than
MIN_CONTOUR_AREA) and ensures they are excluded.

b. Edge Case Testing

 Objective: Handle unexpected inputs gracefully.

 Scenarios:
o No object in the frame: Ensure no errors occur, and the output is blank or
has a message like "No objects detected."
o Multiple overlapping objects: Ensure individual contours are detected and
measured correctly.
o Partially visible objects: Ensure incorrect measurements are flagged or
ignored.

CSE Dept, JCET HUBBALLI Page | 24

 Object Dimension Detection Using Opencv

c. Stress Testing

 Objective: Verify stability under heavy loads.

 Steps:

Use high-resolution video input (e.g., 4K) to ensure the system processes

frames in real-time without crashing.

Introduce complex backgrounds and check if the system can still detect objects
effectively.

d. Usability Testing

 Objective: Ensure the system is user-friendly.

 Steps:
 Test the responsiveness of real-time updates.
o Check for clear and readable dimension overlays on the frame.
o Validate intuitive exit functionality (e.g., pressing q).

6.2 HARDWARE TESTING

a. Camera Testing

 Objective: Ensure the camera captures clear and stable images suitable for
contour detection.
 Steps:
o Test different camera resolutions (e.g., 720p, 1080p, 4K).
o Verify performance under various lighting conditions.
o Check if the camera's frame rate meets real-time requirements (30 FPS or
higher).

CSE Dept, JCET HUBBALLI Page | 25

 Object Dimension Detection Using Opencv

b. System Resource Utilization

 Objective: Ensure the software runs efficiently without overloading the hardware.
 Steps:

o Use system monitoring tools (e.g., Task Manager on Windows or top/htop on

Linux) to measure:

 CPU usage
 Memory consumption
 GPU utilization (if applicable)
o Ensure the system remains responsive while running the application.

c. Hardware Compatibility

 Objective: Verify compatibility with different hardware configurations.

 Steps:
o Test on systems with varying specifications (e.g., different processors,
RAM sizes, and GPUs).
o Verify functionality with different camera models.

CSE Dept, JCET HUBBALLI Page | 26

 Object Dimension Detection Using Opencv

CHAPTER 7

INTERPRETATION OF RESULTS

Fig.7.1 Real Time Dimension Detection Using Camera

Fig.7.2 Real Time Dimension Detection Using Camera

CSE Dept, JCET HUBBALLI Page | 27

 Object Dimension Detection Using Opencv

CONCLUSION

This code captures real-time video, processes frames to detect objects, and calculates their
dimensions using contours and a reference object for calibration. It ensures real-time
visualization of object measurements in centimeters, demonstrating a basic yet effective
system for object dimension detection.

CSE Dept, JCET HUBBALLI Page | 28

 Object Dimension Detection Using Opencv

REFERENCES

1. G. Chandan, Mohana and A. Jain, Real-Time Object Detection and Tracking by

Using OpenCV and Deep Learning, ICIRCA, (2018).
2. W. Wang, Z. Li, J. Deng, H. Yao, X. Xuan, Z. Wang, A review of research on an
object detection based on deep learning, AINIT, (2020).
3. S.Manjula, Dr. K. Lakshmi, A study on object dectection, IJPTFI,ISSN: 0975-
766X, (2016).
4. U. Binay , N. Yadav , Comparative Study of Object Detection Algorithms,
IRJETV4I1103, Volume 4, Issue 7, (2017).
5. Brain Team, V. Quoc Le, Google Research, R. Pang, M. Tan, EffificientDet:
Scalable and Effificient Object Detection, CPR, (2020).
6. Sunil, Gangandeep, Study of object detection methods and applications on
digital images, IJSDR1905088, Volume 4, Issue 5, (2019).

7. K. Shreyamsh, UAV: Application of object detection and tracking techniques

for unmanned aerial vehicles, Texas A & M University, (2015).

CSE Dept, JCET HUBBALLI Page | 29