Optical Character Recognition (OCR) Application For Image To Braille Typeface Conversion
Optical Character Recognition (OCR) Application For Image To Braille Typeface Conversion
2023
[email protected], [email protected]
Mapúa Malayan Colleges Mindanao Davao City, Davao del Sur, Philippines1-5
DOI: https://doi.org/10.54476/ioer-imrj/651032
ABSTRACT
The quantitative research titled "Optical Character Recognition (OCR) Application for Image to
Braille Typeface Conversion" focuses on developing an application capable of converting text-
containing images into a braille PDF document, ready for printing. This initiative aims to benefit
students with visual impairments in the educational environment, offering them a valuable tool for
processing educational materials tailored to their specific needs. The project underwent
evaluation by 10 experts and 20 co-developers, assessing its functionality, responsiveness,
performance, and usability (user satisfaction). Data, collected through Microsoft Forms, were
analyzed using the mean score for each category. The results indicated that Product Functionality,
Product Responsiveness, and Product Performance all received mean scores corresponding to
the interpretation table's "Excellent" category (4.59, 4.52, and 4.51, respectively). Product
Usability (User Satisfaction) achieved a mean score of 4.46, categorized as "Good." Overall, the
application project received an impressive mean score of 4.52, denoting an "Excellent" rating.
This signifies that the application is highly acceptable, effectively fulfilling the intended functions
for end users.
Keywords: Optical Character Recognition (OCR), Braille, Typeface Conversion, Text Extraction,
Application
METHODOLOGY
2. Product Responsiveness
Table 2
Product Responsiveness
Table 3 presents each statement that was
utilized in the process of determining the mean
rating for the performance of the product. Three
items under this category refer to the speed at
which the application processes its various
features. This is the domain where we can also
find the highest mean score for the category,
which is 4.6. This mean score is associated with
the statement, "The speed at which pictures
and/or printed documents can be scanned."
Furthermore, the lowest possible mean score
acquired for this category is 4.3, corresponding to
Each statement used for determining a
the statement, "The speed at which PDF
mean score for the product's responsiveness is
documents can be scanned."
presented in Table 2 below. The statement,
Earning a mean score of 4.51, the
“Capable of notifying the user that they have not
collected data unequivocally indicates that the
detected any text," obtained the highest mean
product's performance aligns with the descriptive
score of 4.6 among all the items. Following "
rating of "excellent." Within this category, three out
Accuracy of text conversion from standard
of five items achieved an "excellent" rating based
typography to braille," with a mean of 4.57. To the
on their mean values, while the remaining items
contrary, the statement, "Capacity to extract each
secured a rating equivalent to "good." This
and every word of the text from the image," has
observation suggests that respondents perceive
the lowest possible mean acquired, a mean score
the application's performance as highly
of 4.40.
satisfactory across all pertinent dimensions,
The aggregated data indicates that the
encompassing processing speed, impact on RAM
product's responsiveness is the descriptive
utilization, and the efficacy of scanning and
equivalent of excellent, with a mean score of 4.52.
conversion processes.
Two of the items in this category obtained the
equivalent of an excellent descriptive score, while
one of them earned a descriptive equivalent of 4. Product Usability
good. This suggests that users see the
application's responsiveness as highly The mean scores received for the various
acceptable, particularly in relation to its ability to aspects of a product's usability, as measured by
alert users of errors, its ability to extract word for user satisfaction, are summarized in Table 4,
word from pictures, and the accuracy of text which can be found below. The data shows that
conversion. the statements, "Satisfaction with the processing
P – ISSN 2651 - 7701 | E – ISSN 2651 – 771X | www.ioer-imrj.com
MAGHANOY, M.S.T., LAPIRAS, M.B., MACASILHIG, B.P.L., IMBIN, N.K.N., NAMION, M.B.B., Optical Character
Recognition (OCR) Application for Image to Braille Typeface Conversion, pp. 146 - 156
150
IOER INTERNATIONAL MULTIDISCIPLINARY RESEARCH JOURNAL, VOL. 5, NO. 3, SEPT. 2023
time" and "Satisfaction with the resulting PDF product's functionality, responsiveness,
document," have the highest mean score of 4.50. performance, and usability (user satisfaction).
The value of 4.37, which corresponds to the item This will serve as the foundation through which
titled "Satisfaction for audio feedback," is the the application's capabilities and usefulness are
category's mean value, which is the lowest evaluated to guarantee conformity with and
possible value. fulfillment of its intended goals.
Table 4 The means and descriptive equivalents for
Usability (User Satisfaction) the five categories of the research assessment
tool used in this study are displayed in the table
provided below. Most of the categories, namely
functionality, responsiveness, and performance,
have a mean score that is consistent with the
description "excellent," as seen by the means
they obtained. While the mean score for the
application's usability was rather close to the good
descriptor. This provides strong evidence to
support the conclusion that the product is highly
acceptable, with a mean score of 4.52.
Considering the foregoing, the image-to-
braille (typeface conversion) application serves a
purpose and is well-received by the respondents.
Thus, the overall data analysis from the
The data presented above show that the research's respondents over alpha testing reveals
category mean for product usability is 4.46, which that the application's functionality,
corresponds to the descriptive equivalent of good. responsiveness, performance, and usability are,
There are four items in this category, two of which as a whole, highly acceptable. As an outcome, it
have a descriptive equivalent of excellent and the is reasonable to presume that the application will
other two of which have a descriptive equivalent fulfill its function for its intended beneficiaries once
of good. This implies that the application can it is made available to the public.
function in accordance with the specifications set
out by the developers, since the respondents have CONCLUSIONS
deemed the application's usability to be
acceptable. This study indicates that the app's
development costs are well-aligned with the goal of
5. Overall Product Rating creating an affordable application for the potential
market. In the alpha testing survey involving 30
Table 5 respondents, the overall mean strongly supports
Overall Product Rating the application's high acceptability, consistent with
the detailed findings in functionality,
responsiveness, performance, and usability. The
Human Activity Assistive Technology (HAAT)
Model (Cook & Hussey, 1995) is invoked to
underscore the importance of considering user
regular activities in addressing their needs, and the
study's positive user satisfaction scores affirm this
approach. Additionally, Goodhue and Thompson's
The general evaluation of the product is (1995) Task-Technology Fit (TTF) Theory
displayed in Table 5. This takes into account the substantiates the excellent mean functionality
Carter, N., Bryant-Lukosius, D., DiCenso, A., Blythe, J., Jackson, L., Powers, P., & Ward, A. (2021). Applying
& Neville, A. J. (2014). The use of triangulation in The HAAT Model To Tackle Equipment
qualitative research. Oncology nursing forum, 41(5), Abandonment. ISSUU.
545–547. https://doi.org/10.1188/14.ONF.545-547 https://issuu.com/nrrts/docs/210281_nrrts_direction
s_issue_1_complete_lr/s/11714561 Johnston, A.
Configure your build. (n.d.). Android Developers. (2017). What is Privacy. IAPP.
https://developer.android.com/studio/build https://iapp.org/about/what-is-privacy/
Cutter, M. & Manduchi R. (2017, October 10). Improving Kahn, S. (2003, April 01).
the accessibility of mobile OCR Apps Via Interactive US6542623B1. United States.
Modalities. National Library of https://worldwide.espacenet.com/patent/search/fami
Medicine.https://pubmed.ncbi.nlm.nih.gov/2927024 ly/023547280/publication/US6542623B1?q=braill
3/ e%20OCR
Davis, F. (1986). Technology acceptance model. Laycock, A., Bailie, J., Matthews, V., & Potvin, L. (2019).
Edutech Wiki. Using developmental evaluation to support
https://edutechwiki.unige.ch/en/Technology_accept knowledge translation: reflections from a large-scale
ance_model quality improvement project in Indigenous primary
healthcare. Health Research Policy and Systems,
Developmental evaluation. (2021, November6). 17(1). https://doi.org/10.1186/s12961-019-0474-6
BetterEvaluation.
https://www.betterevaluation.org/methodsapproach Level Access. (2019, July 02). Understanding assistive
es/approaches/developmental-evaluation technology: How does a blind person use the
internet?
Dilmegani, C. (2020, May 02). Current State of OCR in https://www.levelaccess.com/blog/understanding-
2023: Is it dead or a solved problem? assistive-technology-how- does-a-blind-person-use-
https://research.aimultiple.com/ocr-technology/ the-internet/
Do-It. (2021, April 8). What is braille translation Li, Wang, Doshi, et al. (2018, March 22).
software? https://www.washington.edu/doit/what- US20180082609. United States.
braille-translation-software https://www.freepatentsonline.com/20180082609.p
df
Fareha. (2022, July 07). The biggest problem with OCR
API and how can you fix it?. Filestack. Limitations of using OCR for file classification. (2021).
https://blog.filestack.com/api/biggest-problem-ocr- IDM Magazine. https://idm.net.au/article/0011231-
api-can-fix/. limitations-using-ocr-file-classification/
Gekht N. (2020, February 26). create better backlog and Lokhande et al. (2017, January 04). Braille to text
engage the development team with FURPS. transcription - A literature review. International
GehtSoftUSA. https://gehtsoftusa.com/blog/create- Journal for Scientific Research and
better-backlog-and-engage-the- development-team- Development.
with-furps/
Morton, H., Gunson, N., Marshall, D., McInnes, F., PC’s for students. (2019, June 26). SUNSTAR.
Ayres, A., & Jack, M. (2011). Usability assessment https://www.sunstar.com.ph/article/201129/pcs-for-
of text-to-speech synthesis for additional detail in an students
automated telephone banking system. Computer
Speech & Language, 25(2), 341–362. Preedy, V. R., & Watson, R. R. (2010). 5-point likert
https://doi.org/10.1016/j.csl.2010.05.008 scale. Springer New York EBooks, 4288.
https://doi.org/10.1007/978-0-387-78665-0_6363
Mukherji, S. (2022, March 23). OCR Review: A
comprehensive guide on image to text conversion. Recognize Text in Images with ML Kit on Android.
LinkedIn. https://www.linkedin.com/pulse/ocr- (2023). Firebase.
review-comprehensive-guide-image-text- https://firebase.google.com/docs/ml-
conversion-sujoy-mukherji/?trk=articles_directory kit/android/recognize-text
Nasir, S Z. (2018, June 21). Introduction to arduino Reyes, J. (2022, March 08). How OCR improves content
uno. The Engineering Projects. for physically challenged learners. managed
https://www.theengineeringprojects.com/2018/06/in outsource solutions.
troduction-to-arduino-uno.html
https://www.managedoutsource.com/blog/how-ocr-
Nowell, L. S., Norris, J. M., White, D. E., & Moules, N. J. improves- content-for-physically-challenged-
(2017). Thematic analysis: Striving to meet the learners/.
trustworthiness criteria. International Journal of
Qualitative Methods, 16(1). Rickard, D. M. (n.d.). Vision loss affects our ability to
https://doi.org/10.1177/1609406917733847 communicate. today's
caregiver.https://caregiver.com/articles/vision_loss_
Owayjan, M., Wehbe, T., Daher, E., & Ayoub, O. "The affects_communication/
design and development of a multi-lingual braille
system output device with audio enhancement," Sawant, R., Prabhav, Shrivastava, P., Shahane, P., &
Journal of Software Engineering and Applications, Ramachandran, H. (2021). Text to braille conversion
Vol. 6 No. 5, 2013, pp. 289-295. doi: system. 1-5. 10.1109/ICSES52305.2021.9633940.
10.4236/jsea.2013.65036.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002).
Parienti, R. & Le Guillou F. (2009, September 03). Experimental and quasi-experimental designs for
Portable reading device for the visually impaired. generalized causal inference. Cengage Learning:
WO2009106702A1.France.https://worldwide.espac Boston, MA.
enet.com/patent/search/family/039768902/publicati
Sherman R. (2015). Project management.
on/FR2925202A1?q=WO2009106702A1
science direct.
P – ISSN 2651 - 7701 | E – ISSN 2651 – 771X | www.ioer-imrj.com
MAGHANOY, M.S.T., LAPIRAS, M.B., MACASILHIG, B.P.L., IMBIN, N.K.N., NAMION, M.B.B., Optical Character
Recognition (OCR) Application for Image to Braille Typeface Conversion, pp. 146 - 156
154
IOER INTERNATIONAL MULTIDISCIPLINARY RESEARCH JOURNAL, VOL. 5, NO. 3, SEPT. 2023
https://www.sciencedirect.com/topics/computer- Weisser, R. (2023). ABBYY FineReader helps
science/waterfall-methodology. resources for the blind in the Philippines to produce
braille textbooks for blind students. ABBYY.
Shipman, M. (2022, August 09). Study uncovers how https://www.abbyy.com/customer-stories/abbyy-
blind and visually impaired individuals navigate finereader-helps-resources-for-the-blind-to-
social challenges. NC State University. produce-braille-textbooks/
https://news.ncsu.edu/2022/08/blind-visually-
impaired-challenges/ “What is Alpha Test?.” (2023). Product plan.
https://www.productplan.com/glossary/alphatest/
Shokat, Riaz, Rizvi, et al. (2022, February 25). “What is Beta Test?.” (2023). Product Plan.
Characterization of english braille patterns using https://www.productplan.com/glossary/beta-test/
automated tools and RICA Based Feature Extraction
Methods. Sensors. https://www.mdpi.com/1424- What Is Optical Character Recognition
8220/22/5/1836 (OCR)? (2022, February 18). IBM.
https://www.ibm.com/cloud/blog/optical-character
Susco. (2023). Optical character recognition: how using recognition/
OCR software can increase business efficiency.
https://suscosolutions.com/optical-character- Woodford, C. (2021, May 11). Optical
recognition-using-ocr-software-can- increase- Character Recognition (OCR). Explain that
business-efficiency/ stuff. https://www.explainthatstuff.com/how-ocr-
works.html
Thanki, J. D., Davda, P. D., & Swaminarayan, P. (2021,
April). A Review on OCR Technology. JETIR. Yeom, S. & Kim, D. (2016,
https://www.jetir.org/view?paper=JETIR2104193 March 4). Eye Case.
KR20160024132A. Korea.
Text recognition v2. (2023). https://worldwide.espacenet.com/patent/search/fami
Google developers. ly/055535847/publication/KR20160024132A?q=br
aille%20OCR%20phone
AUTHORS’ PROFILE
https://developers.google.com/ml-kit/vision/text-
recognition/v2 Mica Shaine Maghanoy is an alumnus of
Precious International School of Davao and
Text-to-Speech Technology: What it is and how it works. Mapua Malayan Colleges Mindanao. In 2023, she
(2019, October 16). Reading Rockets. received her high school diploma from the
https://www.readingrockets.org/article/text-speech- institution under the academic track, Science,
technology-what-it-and-how-it-works Technology, Engineering, and Mathematics
(STEM) with 2nd Honors. Currently, she is moving
Timalsina, A. (2023, February 16). Analysis and
forward in committing to a career in healthcare by
benchmarking of OCR accuracy for data extraction presently being a college student enrolled in a
models. DOCSUMO. Bachelor of Science in Nursing.
https://www.docsumo.com/blog/ocr-accuracy