Comparison Between Neural Network and Support Vector Machine in Optical Character Recognition

Phangtriastu, Michael Reynaldo; Harefa, Jeklin; Tanoto, Dian Felita

doi:10.1016/j.procs.2017.10.061

Cited by 41 publications

(18 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A comprehensive comparative study of different classification algorithms applied to Baybayin scripts is also an interesting research direction. Similar to what was shown in Phangtriastu, Harefa & Tanoto (2017), other feature extraction algorithms can be combined with SVM to identify which will work well in classifying Baybayin scripts.…”

Section: Discussionmentioning

confidence: 94%

“…Many script character recognition systems have been reported with SVM as classifiers. In their experiments, Phangtriastu, Harefa & Tanoto (2017) have shown that SVM can achieve an accuracy of 94.43%, and is better when compared to Artificial Neural Network. Tautu & Leon (2012) have studied how SVM can be used to classify handwritten Latin letters, and obtained over 90% precision when tests were implemented on small or capital Latin letters.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optical character recognition system for Baybayin scripts using support vector machine

Pino

Mendoza

Sambayan

2021

PeerJ Computer Science

View full text Add to dashboard Cite

In 2018, the Philippine Congress signed House Bill 1022 declaring the Baybayin script as the Philippines’ national writing system. In this regard, it is highly probable that the Baybayin and Latin scripts would appear in a single document. In this work, we propose a system that discriminates the characters of both scripts. The proposed system considers the normalization of an individual character to identify if it belongs to Baybayin or Latin script and further classify them as to what unit they represent. This gives us four classification problems, namely: (1) Baybayin and Latin script recognition, (2) Baybayin character classification, (3) Latin character classification, and (4) Baybayin diacritical marks classification. To the best of our knowledge, this is the first study that makes use of Support Vector Machine (SVM) for Baybayin script recognition. This work also provides a new dataset for Baybayin, its diacritics, and Latin characters. Classification problems (1) and (4) use binary SVM while (2) and (3) apply the multiclass SVM classification. On average, our numerical experiments yield satisfactory results: (1) has 98.5% accuracy, 98.5% precision, 98.49% recall, and 98.5% F1 Score; (2) has 96.51% accuracy, 95.62% precision, 95.61% recall, and 95.62% F1 Score; (3) has 95.8% accuracy, 95.85% precision, 95.8% recall, and 95.83% F1 Score; and (4) has 100% accuracy, 100% precision, 100% recall, and 100% F1 Score.

show abstract

Section: Discussionmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

Optical character recognition system for Baybayin scripts using support vector machine

Pino

Mendoza

Sambayan

2021

PeerJ Computer Science

View full text Add to dashboard Cite

show abstract

“…(Ghorpade & Katkar, 2014) melakukan penelitian untuk mempelajari algoritma Haar Transform dan Backpropagation Neural Network dengan studi kasus kompresi image, dan penelitian tersebut menghasilkan pernyataan metode Haar Transform dan Backprogaration Neural Network cocok untuk kompresi image dikarenakan menghasilkan image yang memiliki ukuran file kecil tetapi kekurangannya gambar tersebut menjadi pecah pikselnya sehingga untuk studi kasus penelitian ini tidak cocok diterapkan. (Phangtriastu, Harefa, & Tanoto, 2017) melakukan penelitian untuk membanding algoritma Neural Network dengan Support Vector Machine (SVM) menggunakan metode Optical Character Recognition (OCR), dan penelitian ini menghasilkan bahwa metode SVM tidak cocok untuk OCR sehingga diberikan opsi dengan penggabungan antara OCR dan Template Matching dan menghasilkan tingkat akurasi lebih dari 90%. Sehingga metode OCR menjadi landasan yang kuat dalam implementasi pada kasus ini.…”

Section: Pendahuluanunclassified

Identifikasi Plat Nomor Kendaraan Dengan Metode Optical Character Recognition Menggunakan Raspberry Pi

Winarno

Utoro

Prabowo

2020

JI. Jurnal. Informatika

View full text Add to dashboard Cite

Plat Nomor Kendaraan merupakan identitas bagi setiap kendaraan bermotor yang terdaftar oleh pemerintah Indonesia. Proses identifikasi plat nomor diawali dengan pembambilan gambar kendaraan dengan kamera. Gambar akan di resize untuk menyamakan ukuran citra dan crop untuk memisahkan antara plat nomor dengan bagian kendaraan lainnya menggunakan algoritma Perspective Transform, setelah itu di crop kembali untuk memisahkan antara kode wilayah, kode registrasi dan kode seri wilayah, lalu setiap karakter akan dikenali menggunakan metode Optical Character Recognition (OCR) berdasarkan citra karakter hasil crop. Terdapat proses character error handling untuk meningkatkan tingkat akurasi identifikasi pada karakter plat nomor. Plat nomor Indonesia yang diujikan terbagi menjadi 4 kategori yaitu Plat Standar Mobil, Plat Kustom Mobil, Plat Tidak Standar dan Plat Motor. Resolusi kamera terbaik menggunakan resolusi 1280x720 piksel dan menghasilkan rata-rata waktu uji 6,51 detik. Persentase kebenaran identifikasi karakter terbesar untuk plat standar sebesar 100% pada nilai lux 20~39 dan nilai lux 70~99, sedangkan untuk plat nomor kustom persentase kebenaran identifikasi karakter terbesar sebesar 58% pada nilai lux 90~140. Sedangkan untuk plat tidak standar tidak ada persentase kebenaran identifikasi karakter terbesar menghasilkan persentase 0% pada semua nilai lux yang diujikan. Untuk plat nomor motor persentase kebenaran identifikasi karakter terbesar sebesar 8% pada nilai lux 150~199. Sehingga dapat ditarik kesimpulan bahwa plat nomor standar menjadi rekomendasi yang tepat bagi setiap kendaraan bermotor yang ada di Indonesia.

show abstract

“…Other Artificial Neural Network (ANN)-based techniques were used to detect the registration plate area and character segmentation [5,6]. A comparative study of ANNs and SVMs were conducted and results showed that the majority of the experiments are in favor of ANNs [7]. A subset on ANNs is Convolutional Neural Networks (CNNs).…”

Section: Socmentioning

confidence: 99%

Vehicle Classification Through Detection and Color Segmentation of Registration Plates Running on Raspberry Pi 3 Model B

2019

IJRTE

View full text Add to dashboard Cite

Classification of license plates provides useful information regarding the nature of the vehicle, whether it is used for public transport, a privately-owned vehicle, an official vehicle, or a special vehicle. In the Philippines, the registration plates of vehicles are classified by colors. Colors such as red, green blue, black, yellow are used to identify what vehicle classification the plate belongs to. The information is useful to applications in statistics and transport regulation. This paper discusses a convolutional neural network based embedded system that runs on Raspberry Pi 3 Model B. The said system provides a process of classifying vehicles using plate detection using Convolutional Neural Networks and color thresholding of registration plates using the RGB color space. TensorFlow and OpenCV libraries were utilized for the detection and classification.

show abstract

Comparison Between Neural Network and Support Vector Machine in Optical Character Recognition

Cited by 41 publications

References 9 publications

Optical character recognition system for Baybayin scripts using support vector machine

Optical character recognition system for Baybayin scripts using support vector machine

Identifikasi Plat Nomor Kendaraan Dengan Metode Optical Character Recognition Menggunakan Raspberry Pi

Vehicle Classification Through Detection and Color Segmentation of Registration Plates Running on Raspberry Pi 3 Model B

Contact Info

Product

Resources

About