Sistem Pengenal Text berbasis Amazon Web Services

Maulana, Mohammad Reza (2021) Sistem Pengenal Text berbasis Amazon Web Services. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 07111440000143-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2023. Download (1MB) | Request a copy

Abstract

Penyandang tunanetra dikarenakan keterbatasan fisiknya mengalami kesulitan untuk mengetahui sekelilingnya. Mereka membutuhkan alat bantu pembaca peringatan/pesan yang tidak mungkin disajikan dalam huruf Braille. Pada penelitian ini akan dirancang alat bantu dengan cloud-computing platform yang disajikan oleh Amazon Web services (AWS) agar mampu membantu penyandang tunanetra dalam memahami informasi peringatan yang ada pada lingkungan sekelilingnya dengan cepat dan tepat. Adapun dasar pemilihan layanan komputasi awan AWS adalah mengingat bahwa AWS merupakan perusahaan mapan yang sudah beroperasi sejak 2006 dan saat ini merupakan cloud provider terkemuka dan terbaik. Pada penelitian ini jenis layanan yang dimanfaatkan adalah Amazon Textract, yang merupakan layanan Machine Learning pada AWS. Sistem ini berkerja dengan mengenali text dari sebuah citra yang tertangkap oleh kamera bawaan sistem yang kemudian akan dikonversikan menjadi suara ataupun nada peringatan. Sistem ini akan dirancang sebagai Wearable Device atau alat yang dapat disandang oleh pengguna dengan dilengkapi kamera yang dipasangkan ke sistem tertanam (embedded system), dimana akan berkomunikasi dan memanfaatkan layanan atau servis-servis komputasi pada Cloud-provider AWS sehingga mampu mengolah dan mengenali citra-citra yang mengandung teks pesan/peringatan yang ditangkap kamera. Hasil pada penelitian ini adalah Sistem Pengenalan teks yang dapat mengekstrak font resmi dan dirancang sebagai perangkat yang dapat dipakai dan portabel, dimana ketepatan pembacaan mencapai paling rendah 60%.
======================================================================================================
The Visually impaired due to their physical limitations have difficulty on
knowing their surroundings. They need a tool to give them warning by
reading warning signs and written messages that cannot possibly be
presented in braille. In this final project research will design a tool with
a cloud-computing platfrom presented by Amazon Web Services (AWS) to
be able to help the visualy impaired understand on the info of warnings
in their surroundings as quickly and precisely as possible. The basis for
choosing AWS Cloud Computing Services is to remember that Amazon is
an established company that has been operating since 2006 and is
currently the leading and best in cloud providers. In this research, the type
of service used is amazon textract, which is a machine learning service
on AWS. This system works by recognizing the text of an image captured
by the system’s built-in camera which will then be converted into a sound
or warning tonee. This system will be designed as a wearable device or a
device that can be carried by the user with a camera attached to an
embedded system, which will communicate and utilize computing services
or on the AWS Cloud provider so that it is able to process and recognize
the images. It is intended that this wearable aid tool can help the visually
impaired understand their surroundings more quickly, precisely, and
comfortably with the end result of a system that can capture images of
official fonts (Times New Roman, Arial, Calibri, and etc.) which then can
be converted into audible text in quick succession.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Pengenalan Text Image, Sistem Tertanam, Komputasi awan, Layanan Amazon Web Services, Amazon Textract
Subjects:	Q Science > QA Mathematics > QA76.585 Cloud computing. Mobile computing. Q Science > QA Mathematics > QA76.758 Software engineering T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK105.8883 Web authoring software (include web server) T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Mohammad Reza Maulana
Date Deposited:	26 Feb 2021 08:45
Last Modified:	26 Feb 2021 20:58
URI:	http://repository.its.ac.id/id/eprint/82925

Actions (login required)

View Item