Perancangan aplikasi android untuk pengenalan tulisan tangan siswa pada borang jawaban ujian jenis pilihan berganda = Android-based application design for recognizing multiple-choice fill-in handwritten form

Halim, Sandra Puspita (2017) Perancangan aplikasi android untuk pengenalan tulisan tangan siswa pada borang jawaban ujian jenis pilihan berganda = Android-based application design for recognizing multiple-choice fill-in handwritten form. Bachelor thesis, Universitas Pelita Harapan.

[img] Text (Title)
Title.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img]
Preview
Text (Abstract)
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (293kB) | Preview
[img]
Preview
Text (ToC)
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (367kB) | Preview
[img]
Preview
Text (Chapter1)
Chapter1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (251kB) | Preview
[img] Text (Chapter2)
Chapter2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (1MB)
[img] Text (Chapter3)
Chapter3.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (466kB)
[img] Text (Chapter4)
Chapter4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (2MB)
[img] Text (Chapter5)
Chapter5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (285kB)
[img]
Preview
Text (Biblioraphy)
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (290kB) | Preview
[img] Text (Appendices)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (6MB)

Abstract

Ulangan tipe pilihan ganda merupakan salah satu instrumen uji yang populer di kalangan pengajar. Pada instrumen ini, peserta ujian menuliskan jawaban setiap pertanyaan pada borang jawaban yang disediakan. Pemeriksaan borang jawaban dapat dilakukan dengan dua cara, yaitu secara manual dan menggunakan bantuan teknologi Optical Mark Recognition (OMR) atau Optical Character Recognition (OCR). Secara khusus, dengan mengacu pada salah satu hasil penelitian yang memanfaatkan OCR dalam pemeriksaan borang jawaban pilihan ganda, memberikan tingkat akurasi pengenalan tulisan tangan sebesar 69% untuk karakter angka dan 64,5% untuk karakter huruf. Berdasarkan hasil penelitian tersebut, aplikasi dikembangkan dengan pertimbangan bahwa terdapat kemungkinan untuk meningkatkan tingkat akurasi pengenalan tulisan tangan. Aplikasi OCR berbasis Android dalam Tugas Akhir ini difokuskan untuk meningkatkan keakurasian pengenalan karakter. Tahap pra-pengolahan pada aplikasi dimaksud menerapkan pengolahan citra digital. Sedangkan, untuk klasifikasi karakter mengimplementasikan Jaringan Saraf Tiruan (JST) Multilayer Perceptron. Berdasarkan format borang jawaban yang ada, aplikasi melakukan ekstraksi isian pada borang jawaban. Proses ekstraksi dimulai dengan memproses citra borang jawaban yang diambil menggunakan kamera telepon, yang diikuti dengan mengenali tulisan tangan dari citra hasil segmentasi borang jawaban. Hasil pengujian aplikasi menunjukkan bahwa pemrosesan citra borang jawaban yang ditempatkan pada sebidang papan hitam dapat berjalan dengan baik pada saat pencahayaan ruang cukup, jarak pengambilan citra antara 25 cm hingga 28 cm dan kamera membentuk sudut 0o hingga 20o dengan bidang alas borang jawaban, dimana tingkat keberhasilannya dalam memproses 55 sampel citra sebesar 100% untuk deteksi kertas dan 93,3% untuk deteksi kotak borang jawaban. Sementara keberhasilan pengenalan karakter tulisan tangan adalah 84,3% untuk karakter angka, 82,3% untuk nama huruf besar dan 73,9% untuk nama huruf kecil. Pada pengenalan jawaban huruf besar, ‘A’ sampai ‘E’, tingkat keberhasilannya adalah 86,8%. Sedangkan untuk kasus jawaban huruf kecil, ‘a’ sampai ‘e’, keberhasilan klasifikasinya adalah 80.6%. / Multiple choice test is one of the popular test tools among teacher. In this test, student write an answer for each question in the provided answer sheet form. Checking an answer sheet can be done in two way, manually or through the use of technology such as Optical Mark Recognition (OMR) or Optical Character Recognition (OCR). According to a study that use OCR for multiple choice test evaluation, it is concluded that the success rate of handwriting recognition is 69% for digit character and 64.5% for letter character (alphabet). Based on the result, an application is developed with consideration that is possible to increase the accuracy of handwriting recognition. Android-based OCR application in this final project is focused on improving the accuracy of character recognition. The pre-processing phase of the application uses digital image processing. On the other hand, the character classification implements Artificial Neural Network (ANN) Multilayer Perceptron. Based on the form design, the application extracts the content on the image of form. The extraction process starts by processing the image of form taken using a phone camera, followed by handwriting recognition from the result of image segmentation on the form. The result of application test showed that the image processing of answer sheet form that is placed on the black board, can run well when the lighting of the room is sufficient, with distance of taking image at 25 cm to 28 cm and camera form an angle of 0o to 20o. The success rate in processing 55 sample images is 100% to detect paper and 93.3% to detect box answer. Meanwhile, the success rate of handwriting recognition is 84.3% for digit character, 82.3% for uppercase name character, 73.9% for lowercase name character. While for the case of uppercase answer letter, between ‘A’ to ‘E’, the success rate is 86.8% and for lowercase answer letter, between ‘a’ to ‘e’, the success rate is 80.6%.

Item Type: Thesis (Bachelor)
Creators:
CreatorsNIMEmail
Halim, Sandra PuspitaNIM00000004717SANDRAUSPTA@YAHOO.CO.ID
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorMitra, Aditya RamaNIDN0305096901UNSPECIFIED
Thesis advisorKrisnadi, DionNIDN0316029002UNSPECIFIED
Additional Information: SK 82-14 HAL p
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics
Depositing User: Mr Samuel Noya
Date Deposited: 29 Jun 2020 07:53
Last Modified: 01 Sep 2021 08:14
URI: http://repository.uph.edu/id/eprint/8971

Actions (login required)

View Item View Item