Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

A Raspberry Pi-Based Text Reader &

Object Detection System
Kiran Gurav1; Neel Joshi2; Nida Desai3; Shruti Ghorpade4
Kolhapur Institute of Technology’s College of Engineering,
Kolhapur

Abstract:- This research introduces a novel application of I. INTRODUCTION

the raspberry pi platform in the development of an
integrated text reader and object detection system. The This research presents a Raspberry Pi-based system
system aims to assist visually impaired individuals and designed to assist visually impaired individuals. The system
enhance overall accessibility for users with diverse needs. combines text reading and object detection capabilities. It
Leveraging the power of computer vision and edge employs OCR to convert printed or handwritten text into
computing, the proposed solution employs a raspberry pi, audio, and object detection to identify and describe
a low-cost, compact, and energy-efficient single-board surrounding objects. The goal is to enhance independence and
computer. safety through auditory feedback. The system is cost-
effective, portable, and shows promise in improving the lives
The text reader component utilizes optical character of visually impaired users. The system combines two key
recognition (ocr) algorithms to convert printed or functionalities: a text reader and an object detection module.
handwritten text into digital format. This feature enables
users to obtain audible information from text-based The text reader component utilizes Optical Character
sources, such as books, documents, or signage. The Recognition (OCR) technology to convert printed or
implementation of real-time processing on the raspberry handwritten text into digital format. This allows users to
pi ensures prompt and on-device text recognition, access information from text-based sources like books,
reducing dependence on external servers and enhancing documents, or signage through real-time audio feedback.
privacy.
Furthermore, the object detection module incorporates a
Furthermore, the system incorporates an object pre-trained deep neural network to identify and describe
detection module to identify and describe objects in the objects in the user's surroundings. This enhances spatial
user's surroundings. This functionality enhances the awareness by providing audio cues about the presence and
user's spatial awareness by providing audio cues about characteristics of objects, fostering greater independence and
the presence and characteristics of objects. The object safety.
detection system employs a pre-trained deep neural
network, making it adaptable to various object II. REVIEW OF RELATED LITERATURE
recognition tasks.
The review of related literature for a raspberry pi-based
The entire system is designed with a user-friendly text reader & object detection system encompasses various
interface that facilitates interaction through speech and aspects, including the utilization of raspberry pi in assistive
audio feedback. Additionally, the portability and technologies, advancements in optical character recognition
affordability of the raspberry pi make the solution (ocr) and object detection, and the significance of such
accessible to a broad user base. systems for individuals with visual impairments. Here is an
overview of the key literature areas:
Preliminary testing of the prototype has shown
promising results in terms of accuracy, speed, and A. Raspberry pi in Assistive Technologies:
usability. The raspberry pi-based text reader & object numerous studies have explored the applications of
detection system holds potential for improving the quality raspberry pi in diverse fields, including healthcare and
of life for individuals with visual impairments and can assistive technologies. Researchers highlight the platform's
contribute to the advancement of assistive technologies versatility, low cost, and energy efficiency, making it an
with its cost-effective and scalable approach. Future work attractive choice for developing solutions that cater to specific
involves refining the system, expanding the object needs, such as aiding individuals with visual impairments
detection capabilities, and conducting extensive user trials (molleman et al., 2017).
to gather feedback for further improvements.

IJISRT24AUG491 www.ijisrt.com 1108

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

B. Optical Character Recognition (OCR) Advances: evaluation is identified as lacking comprehensive

literature on ocr technologies reveals ongoing assessments of the proposed assistive technology's usability,
advancements in improving accuracy and speed. Researchers accessibility, and impact on visually impaired users.
have focused on optimizing ocr algorithms for real-time Scalability and generalization represent a gap, as research
processing, and some studies emphasize the importance of often involves specific models, emphasizing the need to
implementing ocr on edge devices to enhance accessibility explore techniques for adaptability across diverse contexts.
and privacy (liang et al., 2018; luo et al., 2020).
Energy efficiency and power consumption is recognized
C. Object detection and recognition: as a gap, urging research to quantify the energy consumption
Object detection has seen significant progress with the of the proposed system during various tasks to optimize
rise of deep learning techniques, particularly convolutional power usage without compromising performance. Long-term
neural networks (cnns). Research in this area discusses the user adoption and impact is highlighted as an underexplored
development of robust and efficient object detection models, area, calling for research on how users integrate the system
which can be crucial for creating reliable assistive into their daily routines over an extended period, considering
technologies (redmon et al., 2018; liu et al., 2016). factors such as updates, maintenance, and user adaptation.
Addressing these gaps can refine the raspberry pi-based text
D. Assistive Technologies for the Visually Impaired: reader & object detection system, ensuring practicality, user
there is a rich body of literature highlighting the acceptance, and positive impact on visually impaired
importance of assistive technologies for individuals with individuals.
visual impairments. Studies emphasize the positive impact of
technologies that provide auditory feedback, navigation The existing system in market nowadays for visually
assistance, and object recognition for improving the impaired people mostly focuses on object detection.
independence and daily lives of visually impaired individuals Meanwhile, there is no cost friendly solution for object
(coughlan et al., 2017; pradhan et al., 2018). recognition with speech conversion module available.

E. User Interface and Experience in Assistive Technologies:  Objective:

Research on user interface design and user experience Our objective is to design the project with low cost and
in assistive technologies is critical for ensuring the all level users can have for object recognition. Hence this
effectiveness and acceptance of such systems. Human- project seeks to design a object recognition system that will
computer interaction studies guide the development of use a yolo (you look only once) algorithm and hence real-time
intuitive interfaces that cater to the specific needs and objects in an image are identified and converted to their name
preferences of users with visual impairments (andré et al., and these names are converted to speech signals.
2018; brewster et al., 2019).
 Scope
In summary, the existing literature provides a solid Object detection is a cutting-edge technology that has
foundation for the development of a raspberry pi-based text the potential to improve many Aspects of our lives. So, one
reader & object detection system, emphasizing the platform's of its applications is object recognition. However, it’s
potential, advancements in ocr and object detection, and the Important to note that this technology is not a one-size-fits-
broader context of assistive technologies for individuals with all solution, and the
visual impairments. The synthesis of these findings informs Choice of a specific method will depend on the application
the design, implementation, and evaluation of the proposed and available data.
system.
Furthermore, ethical and legal considerations need to be
III. RESEARCH GAP taken into account when Using this is privacy and data
security. One of the key areas of focus for researchers is to
The literature review reveals insights into the use of improve the real-time performance Of its methods. As the
raspberry pi in assistive technologies, OCR, and object technology becomes more powerful, it can detect and track
detection for the visually impaired. However, there are Objects up to date, making it more useful for many
notable research gaps. Integration of ocr and object detection applications. Additionally, Researchers are working on
is an area where studies often discuss these technologies developing more robust and accurate algorithms that can
separately, missing an opportunity to explore their seamless Better handle different object scales, orientations, and
combination in real-time scenarios for enhanced usability. lighting conditions. Another Area of focus is the integration
Performance optimization on edge devices is identified as a of object detection with other technologies, such as
gap, necessitating detailed investigations into tailoring ocr Augmented reality.
and object detection algorithms for resource-constrained
devices like raspberry pi to improve speed and accuracy. Besides, there will be a continued emphasis on making
this technology more Accessible to a broader range of users.
The adaptability to dynamic environments is another This will involve the development of more User-friendly
research gap, with a focus on static scenarios rather than tools and interfaces, as well as easily fine-tuned pre-trained
addressing challenges in dynamic settings. User-centric models For specific applications.

IJISRT24AUG491 www.ijisrt.com 1109

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

IV. LIST OF MATERIALS

A. Hardware Component:

 Raspberry Pi 4 Model B:

Fig 1 Raspberry pi with it’s port name used as controller

Raspberry pi 4 model b is the latest product in the  Speaker:

popular raspberry pi range of computers. It offers ground-
breaking increases in processor speed, multimedia
performance, memory, and connectivity compared to the
prior-generation raspberry pi 3 model b+, while retaining
backwards compatibility and similar power consumption. For
the end user, raspberry pi 4 model b provides desktop
performance comparable to entry-level x86 pc systems. This
product’s key features include a high-performance 64-bit
quad-core processor, dual-display support at resolutions up to
4k via a pair of micro-hdmi ports, hardware video decode at
up to 4kp60, up to 8gb of ram, dual-band 2.4/5.0 ghz wireless
lan, bluetooth 5.0, gigabit ethernet, usb 3.0, and poe
capability (via a separate poe hat add-on). The dual-band
wireless lan and bluetooth have modular compliance
certification, allowing the board to be designed into end
products with significantly reduced compliance testing,
improving both cost and time to market 15
Fig 2: Speaker to Listen Output

 Output power: 3wx2

 Battery voltage: 3.7v
 Input power supply: 5v
 Impedance is around 4 ohms

IJISRT24AUG491 www.ijisrt.com 1110

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

 Web Camera:

Fig 3: Web Cam Used to Take Input

A webcam is a video camera which is designed to record This is the official operating system developed for
or stream to a computer or computer network. They are raspberry pi by the raspberry pi foundation. It is a debian-
primarily used in video telephony, live streaming and social based Linux distribution optimized for the pi's hardware and
media, and security. In this project webcam is used to capture comes with a user-friendly interface.
the object and send its image further into the system.
 Code:
B. Software Components: https://drive.google.com/drive/folders/1rAa67INfWBox
FePBdj4qifZhF85qm1rt?usp=sharing
 Raspberry pi os (formerly raspbian) Please send a mail before using code.

Fig 4: Raspberry Pi OS is the official operating system for

Raspberry Pi devices

IJISRT24AUG491 www.ijisrt.com 1111

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

Fig 5: Project Code

(Mail us while requesting for access at B. Hardware Setup:

kiran22gurav@gmail.com)
 Acquire Hardware:
(change location of coco.name with respect to your
computer location path. Same for other library path. To  obtain a raspberry pi board (e.g., raspberry pi 3 or 4).
download library just search respective libraries name on  connect essential peripherals, including a microsd card,
google. power supply, keyboard, mouse, and monitor.

Thank you.)  Camera Setup:

V. PROCEDURE  If using a camera module (e.g., raspberry pi camera

module), connect it to the designated camera port on the
A. Mount the Hardware as Per Below Figure. raspberry pi.
Creating a raspberry pi-based text reader & object
detection system involves several steps, including hardware C. Software Installation:
setup, software installation, programming, and integration.
Below is a general procedure to guide you through the  Install raspberry Pi Os:
development process. Please note that this is a high-level
overview, and you may need to refer to specific  Download the latest version of raspberry pi os (formerly
documentation for detailed instructions on each step. raspbian) and install it on the microsd card using tools like
etcher.
 Boot up the raspberry pi with the microsd card.

IJISRT24AUG491 www.ijisrt.com 1112

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

 Update System Packages: D. Text Reader Implementation:

 open a terminal and run:  Text Recognition Setup:

 sudo apt update
 sudo apt upgrade  install ocr software (e.g., tesseract):
 sudo apt install tesseract-ocr
 Install Required Libraries:
 Text-to-Speech (TTS) Setup:
 install necessary libraries for the camera, gpio, and other
components:  choose a tts engine (e.g., espeak) and install it:
 sudo apt install python3-picamera  sudo apt install espeak

Fig 6: Proteus Implementation and Running

(File Uploaded on Drive Link)

 Python Programming:  Download Pre-Trained Model:

 write a python script to capture images using the camera,  choose an object detection model compatible with
perform ocr on the images, and convert the recognized tensorflow lite (e.g., mobilenet ssd).
text to speech.  download the model and labels file.
 utilize libraries like opencv and pytesseract.
 Tensorflow Lite Installation:

E. Object Detection Implementation:  install tensorflow lite:

 pip install tflite-runtime
 Object Detection Setup:
 Python Programming for Object Detection:
 install opencv for object detection:
 pip install opencv-python  write a python script to capture images, perform object
detection using the pre-trained model, and display the
results.
 use opencv and tensorflow lite for implementation.

IJISRT24AUG491 www.ijisrt.com 1113

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

F. Integration:

 Combine Text Reader and Object Detection:

 create a unified python script that incorporates both the

text reading and object detection functionalities.
 ensure efficient resource usage to prevent performance
issues on the raspberry pi.

 Testing:

 test the system with sample images and text to verify the
accuracy of both the text reading and object detection
components.
 refine the script as needed.

G. Optional Enhancements:

 User Interface (UI):

 Create a simple ui for user interaction, using tools like

tkinter.

 Logging and Analytics:

 Implement logging to record system events, errors, and

performance metrics.
 Use data analysis tools for further insights.

 Optimization:

 Optimize the code for performance, considering the

limited resources of the raspberry pi.

 Documentation:

 document the system architecture, software dependencies,

and usage instructions for future reference.
Fig 7: Data Flow Diagram
By following this procedure and customizing it based on
your specific requirements and components, you can develop VI. DATA ANALYSIS TOOLS
a raspberry pi-based text reader & object detection system.
 Python (with Pandas and Numpy):

 Description: python is well-suited for the raspberry pi

environment. Pandas and numpy can handle data
manipulation and basic statistical analysis efficiently.
Python's versatility also makes it suitable for scripting and
automation.

 Opencv:

 Description: opencv (open source computer vision

library) is essential for image processing and computer
vision tasks. It can be utilized for analyzing images and
video streams, extracting features, and performing object
detection.

IJISRT24AUG491 www.ijisrt.com 1114

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

 Tensorflow Lite: B. Accessibility and inclusivity:

Emphasize the significance of the text reader component
 Description: tensorflow lite is a lightweight version of the in promoting accessibility for individuals with visual
popular machine learning framework. It is suitable for impairments. Discuss how the system contributes to fostering
running trained models on resource-constrained devices inclusivity by converting visual information into auditory
like the raspberry pi, making it ideal for object detection output, thereby empowering users who rely on assistive
tasks. technologies.

 Matplotlib and Seaborn: C. Object Detection Applications:

Explore the various practical applications of the object
 Description: matplotlib and seaborn can be used for detection system. Discuss how it can be deployed in different
creating simple visualizations directly on the raspberry pi. contexts such as security, inventory management, and
They are lightweight and can generate plots and charts to beyond. Evaluate the system's ability to accurately identify
aid in the analysis of data. and classify objects, and consider potential enhancements to
expand its versatility.
 Scikit-Learn:
D. Integration challenges and solutions:
 Description: scikit-learn provides simple and efficient Address any integration challenges encountered when
tools for data mining and data analysis. It includes various combining the text reader and object detection functionalities.
algorithms for classification, regression, clustering, and Discuss how the raspberry pi platform facilitated or posed
dimensionality reduction, which can be beneficial for limitations to integration. Consider the modularity of the
further analysis of the system's performance. system and how easily it can be adapted for diverse use cases.

 Sqlite: E. Cost-Effectiveness and Accessibility:

Highlight the cost-effectiveness of the raspberry pi
 Description: sqlite is a lightweight and embedded platform and its role in making advanced technologies more
relational database management system. It can be accessible. Discuss how this affordability contributes to the
employed for local storage and retrieval of structured data, democratization of technology, enabling a broader range of
such as logs, configurations, or metadata related to the text individuals and organizations to implement similar systems.
reader and object detection system.
F. Open-Source Community Collaboration:
 Jupyter Notebooks (On A Remote Server): Discuss the benefits of the open-source nature of the
raspberry pi community. Explore how collaboration within
 Description: while running jupyter notebooks directly on this community can lead to continuous improvement,
a raspberry pi might be resource-intensive, you can updates, and the development of additional features. Consider
consider running them on a more powerful server and the potential for community-driven enhancements and
accessing them remotely from the raspberry pi. This contributions.
allows you to perform interactive and exploratory data
analysis. G. Limitations and Future Directions:
Acknowledge any limitations of the current system and
 Bash and Shell Scripting: propose potential avenues for future research and
development. Discuss areas where improvements could be
 Description: for basic log analysis, automation, and made, such as refining algorithms for better accuracy,
system monitoring, bash and shell scripting on the expanding language support, or incorporating additional
raspberry pi can be valuable. These scripts can help parse sensors for enhanced functionality.
logs, monitor system performance, and execute routine
tasks. H. Ethical Considerations:
Consider ethical implications related to privacy and data
VII. DISCUSSION security, especially in the context of object detection
applications. Discuss any measures taken to address these
A. Performance Evaluation: concerns and emphasize the importance of responsible
Discuss the performance of the text reader and object development and deployment of such systems.
detection components individually. Highlight any challenges
faced during development and implementation, and provide VIII. CONCLUSION
insights into how these challenges were addressed. Evaluate
the accuracy, speed, and overall effectiveness of the system We have implemented an image to speech conversion
in real-world scenarios. technique using a raspberry pi. The simulation results have
been successfully verified and the hardware output has been
tested using different samples. Our algorithm successfully
processes the image and reads it out clearly. This is an

IJISRT24AUG491 www.ijisrt.com 1115

Volume 9, Issue 8, August – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24AUG491

economical as well as an efficient device for the visually REFERENCES

impaired people. By implementing this method, they will get
a chance to overcome the problems faced by impaired in the [1]. Wikipedia raspberry pi:
present. This technique act as a third eye for the blind people. https://en.wikipedia.org/wiki/raspberry_pi
We have applied our algorithm to many images of object and [2]. S. Koley and r. Mishra voice operated outdoor
found that it successfully does its conversion. Even many navigation system for visually impaired persons
such applications can be made for benefit of society and this international journal of engineering trends and
device is compact and helpful to the society. technology, vol.3, issue 2,2012.
[3]. S. Dhambre and a. Sakare smart stick for blind:
In conclusion, the development and implementation of obstacle detection, artificial vision and real-time
a raspberry pi-based text reader and object detection system assistance via gps,”2nd international conference on
represent a significant stride towards enhancing accessibility information and communication technology
and automation in various domains. Through the integration (ncict)2011.
of cutting-edge technologies, such as computer vision and [4]. Ajantha devi and santhosh baboo`s “embedded optical
natural language processing, this research has successfully character recognition on tamil text image using
demonstrated the feasibility and practicality of utilizing a raspberry pi”, international journal of computer
cost-effective and versatile platform like the raspberry pi for science trends and technology (ijcst), volume. 2, issue
these purposes. 4, july-august 2014.
[5]. Vasanthi. G and ramesh babu y india. Vision based
The text reader component of the system showcases the assistive system for label detection with voice output,
potential for empowering individuals with visual department of ece, dmi college of engineering,
impairments by providing real-time text-to-speech chennai, jan 2014.
capabilities. By leveraging optical character recognition (ocr) [6]. Vikram shirol, abhijit m, savitri a et al. “drashti an
algorithms and synthesizing speech output, the system offers android reading aid” international journal of computer
a valuable tool for promoting inclusivity and independence. science and information technologies vol.6 (july 2015)
Chatgpt for code
Furthermore, the object detection module extends the [7]. Raspberry pi os
system's utility to broader applications, spanning from https://www.raspberrypi.com/software/operating-
security surveillance to inventory management. The ability to systems/
identify and classify objects in real-time not only enhances [8]. Open-cv download https://opencv.org/releases/
operational efficiency but also lays the foundation for the
integration of intelligent systems into various aspects of our
daily lives.

The open-source nature of the raspberry pi community

facilitates continuous improvement and customization of the
system, encouraging collaboration and innovation. Future
work may involve refining the accuracy of object detection
algorithms, expanding the language support of the text reader,
and exploring opportunities for additional functionalities.

In summary, the raspberry pi-based text reader and

object detection system presented in this research paper
represent a commendable step towards democratizing
advanced technologies. As we continue to harness the
potential of affordable and accessible platforms, the impact
on diverse fields is expected to grow, fostering a more
inclusive and automated future.

IJISRT24AUG491 www.ijisrt.com 1116