Garuda - Garba Rujukan Digital

Perbandingan Kinerja Support Vector Machine (SVM) Dalam Mengenali Wajah Menggunakan SURF DAN GLCM Bahri, Syamsul; Saddami, Khairun; Arnia, Fitri; Muchtar, Kahlil
JURNAL NASIONAL TEKNIK ELEKTRO Vol 8, No 2: July 2019
Publisher : Jurusan Teknik Elektro Universitas Andalas

Face recognition is one part of the biometrics research. Face recognition is widely used in identification and recognition process. Speed-up Robust Feature (SURF) is one of feature extraction method used in face recognition system. This research aims to compare face recognition performance between SURF and Gray Level Co-occurence Matrix (GLCM) methods for perspective rotation. In this study, the image features were extracted using SURF and GLCM. Each feature was used on classification stage using Support Vector Machine (SVM). The dataset was obtained from National Cheng Kung University (NCKU). The NCKU dataset has more variation of rotation angle. The dataset used in this study consists of 10 classes that showed 10 of the subject. The results show that SURF method obtained 85% of accuracy and GLCM method reached 50% of accuracy. Therefore, we concluded that SURF method has better performance on implementing on face recognition system.Keywords : SURF, GLCM, Face Recognition, SVMÂ AbstrakÂ Pengenalan wajah merupakan salah satu bagian dari penelitian biometrika. Pengenalan wajah banyak digunakan dalam proses identifikasi manusia. Metode ekstraksi fitur Speed-Up Robust Feature (SURF) merupakan salah satu metode yang digunakan untuk mengenali wajah. Penelitian ini bertujuan untuk membandingkan kinerja sistem pengenalan wajah dengan menggunakan metode ekstraksi fitur SURF dan Gray Level Co-occurence Matrix (GLCM). Pada penelitian ini, data input wajah akan diekstraksi fiturnya menggunakan SURF dan GLCM. Setiap fitur digunakan pada tahapan klasifikasi menggunakan Support Vector Machine (SVM). Data yang digunakan merupakan data yang didapatkan dari National Cheng Kung University (NCKU). Data wajah NCKU mempunyai sudut rotasi yang lebih banyak. Dataset yang digunakan pada penelitian ini terdiri dari 10 kelas yang menunjukkan 10 subjek penelitian. Pengenalan wajah menggunakan metode SURF dan SVM mempunyai akurasi 85%, sedangkan menggunakan metode GLCM mempunyai akurasi 50%. Hasil menunjukkan bahwa metode SURF mempunyai kinerja yang lebih baik dari metode GLCM.Kata Kunci : SURF, GLCM, pengenalan wajah, SVM

Improvement of binarization performance using local otsu thresholding Khairun Saddami; Khairul Munadi; Yuwaldi Away; Fitri Arnia
International Journal of Electrical and Computer Engineering (IJECE) Vol 9, No 1: February 2019
Publisher : Institute of Advanced Engineering and Science

Ancient document usually contains multiple noises such as uneven-background, show-through, water-spilling, spots, and blur text. The noise will affect the binarization process. Binarization is an extremely important process in image processing, especially for character recognition. This paper presents an improvement to Nina binarization technique. Improvements were achieved by reducing processing steps and replacing median filtering by Wiener filtering. First, the document background was approximated by using Wiener filter, and then image subtraction was applied. Furthermore, the manuscript contrast was adjusted by mapping intensity of image value using intensity transformation method. Next, the local Otsu thresholding was applied. For removing spotting noise, we applied labeled connected component. The proposed method had been testing on H-DIBCO 2014 and degraded Jawi handwritten ancient documents. It performed better regarding recall and precision values, as compared to Otsu, Niblack, Sauvola, Lu, Su, and Nina, especially in the documents with show-through, water-spilling and combination noises.

Moment invariant-based features for Jawi character recognition Fitri Arnia; Khairun Saddami; Khairul Munadi
International Journal of Electrical and Computer Engineering (IJECE) Vol 9, No 3: June 2019
Publisher : Institute of Advanced Engineering and Science

Ancient manuscripts written in Malay-Arabic characters, which are known as "Jawi" characters, are mostly found in Malay world. Nowadays, many of the manuscripts have been digitalized. Unlike Roman letters, there is no optical character recognition (OCR) software for Jawi characters. This article proposes a new algorithm for Jawi character recognition based on Hu’s moment as an invariant feature that we call the tree root (TR) algorithm. The TR algorithm allows every Jawi character to have a unique combination of moment. Seven values of the Hu’s moment are calculated from all Jawi characters, which consist of 36 isolated, 27 initial, 27 middle, and 35 end characters; this makes a total of 125 characters. The TR algorithm was then applied to recognize these characters. To assess the TR algorithm, five characters that had been rotated to 90o and 180o and scaled with factors of 0.5 and 2 were used. Overall, the recognition rate of the TR algorithm was 90.4%; 113 out of 125 characters have a unique combination of moment values, while testing on rotated and scaled characters achieved 82.14% recognition rate. The proposed method showed a superior performance compared with the Support Vector Machine and Euclidian Distance as classifier.

Perbandingan Kinerja Support Vector Machine (SVM) Dalam Mengenali Wajah Menggunakan SURF DAN GLCM Syamsul Bahri; Khairun Saddami; Fitri Arnia; Kahlil Muchtar
JURNAL NASIONAL TEKNIK ELEKTRO Vol 8, No 2: July 2019
Publisher : Jurusan Teknik Elektro Universitas Andalas

Face recognition is one part of the biometrics research. Face recognition is widely used in identification and recognition process. Speed-up Robust Feature (SURF) is one of feature extraction method used in face recognition system. This research aims to compare face recognition performance between SURF and Gray Level Co-occurence Matrix (GLCM) methods for perspective rotation. In this study, the image features were extracted using SURF and GLCM. Each feature was used on classification stage using Support Vector Machine (SVM). The dataset was obtained from National Cheng Kung University (NCKU). The NCKU dataset has more variation of rotation angle. The dataset used in this study consists of 10 classes that showed 10 of the subject. The results show that SURF method obtained 85% of accuracy and GLCM method reached 50% of accuracy. Therefore, we concluded that SURF method has better performance on implementing on face recognition system.Keywords : SURF, GLCM, Face Recognition, SVM Abstrak Pengenalan wajah merupakan salah satu bagian dari penelitian biometrika. Pengenalan wajah banyak digunakan dalam proses identifikasi manusia. Metode ekstraksi fitur Speed-Up Robust Feature (SURF) merupakan salah satu metode yang digunakan untuk mengenali wajah. Penelitian ini bertujuan untuk membandingkan kinerja sistem pengenalan wajah dengan menggunakan metode ekstraksi fitur SURF dan Gray Level Co-occurence Matrix (GLCM). Pada penelitian ini, data input wajah akan diekstraksi fiturnya menggunakan SURF dan GLCM. Setiap fitur digunakan pada tahapan klasifikasi menggunakan Support Vector Machine (SVM). Data yang digunakan merupakan data yang didapatkan dari National Cheng Kung University (NCKU). Data wajah NCKU mempunyai sudut rotasi yang lebih banyak. Dataset yang digunakan pada penelitian ini terdiri dari 10 kelas yang menunjukkan 10 subjek penelitian. Pengenalan wajah menggunakan metode SURF dan SVM mempunyai akurasi 85%, sedangkan menggunakan metode GLCM mempunyai akurasi 50%. Hasil menunjukkan bahwa metode SURF mempunyai kinerja yang lebih baik dari metode GLCM.Kata Kunci : SURF, GLCM, pengenalan wajah, SVM

Kombinasi Metode Nilai Ambang Lokal dan Global untuk Restorasi Dokumen Jawi Kuno Khairun Saddami; Fitri Arnia; Yuwaldi Away; Khairul Munadi
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 7 No 1: Februari 2020
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2020701741

Dokumen Jawi kuno merupakan warisan budaya yang berisi informasi penting tentang peradaban masa lalu yang dapat dijadikan pedoman untuk masa sekarang ini. Dokumen Jawi kuno telah mengalami penurunan kualitas yang disebabkan oleh beberapa faktor seperti kualitas kertas atau karena proses penyimpanan. Penurunan kualitas ini menyebabkan informasi yang terdapat pada dokumen tersebut menghilang dan sulit untuk diakses. Artikel ini mengusulkan metode binerisasi untuk membangkitkan kembali informasi yang terdapat pada dokumen Jawi kuno. Metode usulan merupakan kombinasi antara metode binerisasi berbasis nilai ambang lokal dan global. Metode usulan diuji terhadap dokumen Jawi kuno dan dokumen uji standar yang dikenal dengan nama Handwritten Document Image Binarization Contest (HDIBCO) 2016. Citra hasil binerisasi dievaluasi menggunakan metode: F-measure, pseudo F-measure, peak signal-to-noise ratio, distance reciprocal distortion, dan misclasification penalty metric. Secara rata-rata, nilai evaluasi F-measure dari metode usulan mencapai 88,18 dan 89,04 masing-masing untuk dataset Jawi dan HDIBCO-2016. Hasil ini lebih baik dari metode pembanding yang menunjukkan bahwa metode usulan berhasil meningkatkan kinerja metode binerisasi untuk dataset Jawi dan HDIBCO-2016. AbstractAncient Jawi document is a cultural heritage, which contains knowledge of past civilization for developing a better future. Ancient Jawi document suffers from severe degradation due to some factors such as paper quality or poor retention process. The degradation reduces information on the document and thus the information is difficult to access. This paper proposed a binarization method for restoring the information from degraded ancient Jawi document. The proposed method combined a local and global thresholding method for extracting the text from the background. The experiment was conducted on ancient Jawi document and Handwritten Document Image Binarization Contest (HDIBCO) 2016 datasets. The result was evaluated using F-measure, pseudo F-measure, peak signal-to-noise ratio, distance reciprocal distortion, dan misclassification penalty metric. The average result showed that the proposed method achieved 88.18 and 89.04 of F-measure, for Jawi and HDIBCO-2016, respectively. The proposed method resulted in better performance compared with several benchmarking methods. It can be concluded that the proposed method succeeded to enhance binarization performance.

On Reducing ShuffleNets’ Block for Mobile-based Breast Cancer Detection Using Thermogram: Performance Evaluation Rizka Ramadhana; Khairun Saddami; Khairul Munadi; Fitri Arnia
Indonesian Journal of Electrical Engineering and Informatics (IJEEI) Vol 10, No 4: December 2022
Publisher : IAES Indonesian Section

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52549/ijeei.v10i4.4062

In this paper, we proposed a reduced-block-Shufflenet (RB-ShuffleNet) for thermal breast cancer detection. RB-ShuffleNet is a modification of Shufflenet obtained by reducing blocks from the original architecture. The images for training and testing were obtained from Database for Mastology Research (DMR). First, we detected and cropped the image based on the region of interest (ROI), in which the ROI is determined by using the red intensity profile. Then, the ROI images were trained using RB-ShuffleNets. In the experiments, we built eight architectures, based on ShuffleNet, each with a different number of reduced blocks. The result showed that RB-Shufflenet with four reduced blocks had fewer than 50% of the learning parameters of the original Shufflenet, without compromising its performance. The RB-ShuffleNet with up to four reduced blocks could achieve 100% testing accuracy. Furthermore, The RB-ShuffleNets performed better than MobileNetV2 and resulted in higher accuracy when fed with ROI images. Due to its light structure and good performance, we recommend RB-ShuffleNet as mobile-based CNN model which is preferable to implement in breast cancer detection.

Prediksi Kecepatan Angin Jangka Menengah dengan Artificial Neural Network untuk Estimasi Daya Listrik Tenaga Angin (Studi Kasus: Kota Sabang) Abdul Malek; Suriadi Suriadi; Khairun Saddami
Jurnal Serambi Engineering Vol 8, No 3 (2023): Juli 2023
Publisher : Fakultas Teknik

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/jse.v8i3.6010

Indonesia, as a country at the equator, has a very large renewable energy potential that can be used as a source of electrical energy. Electricity consumption in Indonesia, especially in Aceh, continues to increase annually because of population growth and increasing economic needs. Recently, the construction of power plants has been considered to be environmentally friendly and economical. One of the efforts that can be made is the development of wind-power plants. The availability of certain wind speeds was expected. Therefore, accurate prediction data must be used as the basis for building wind power plants. To increase the accuracy of wind speed prediction by looking at the error rate in predicting the amount of wind speed generated using an Artificial Neural Network with feed-forward and feed-backward functions from the back propagation algorithm (BPNN). The results of the application using the Neural Network algorithm with a back propagation Neural Network (BPNN) to predict wind speed show that the Neural Network algorithm can predict wind speed with an error of 0.0036. In July 2021, the estimated energy demand is 81.5 KWH.

Prediksi Kecepatan Angin Jangka Menengah dengan Artificial Neural Network untuk Estimasi Daya Listrik Tenaga Angin (Studi Kasus: Kota Sabang) Abdul Malek; Suriadi Suriadi; Khairun Saddami
Jurnal Serambi Engineering Vol 8, No 3 (2023): Juli 2023
Publisher : Fakultas Teknik

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32672/jse.v8i3.6010

Indonesia, as a country at the equator, has a very large renewable energy potential that can be used as a source of electrical energy. Electricity consumption in Indonesia, especially in Aceh, continues to increase annually because of population growth and increasing economic needs. Recently, the construction of power plants has been considered to be environmentally friendly and economical. One of the efforts that can be made is the development of wind-power plants. The availability of certain wind speeds was expected. Therefore, accurate prediction data must be used as the basis for building wind power plants. To increase the accuracy of wind speed prediction by looking at the error rate in predicting the amount of wind speed generated using an Artificial Neural Network with feed-forward and feed-backward functions from the back propagation algorithm (BPNN). The results of the application using the Neural Network algorithm with a back propagation Neural Network (BPNN) to predict wind speed show that the Neural Network algorithm can predict wind speed with an error of 0.0036. In July 2021, the estimated energy demand is 81.5 KWH.

Water Level Detection for Flood Disaster Management Based on Real-time Color Object Detection Saddami, Khairun; Nurdin, Yudha; Noviantika, Fina; Oktiana, Maulisa; Muchallil, Sayed
Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control Vol. 8, No. 1, February 2023
Publisher : Universitas Muhammadiyah Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22219/kinetik.v8i1.1635

Currently, the water level monitoring system for a river uses instruments installed on the banks of the river and must be checked continuously and manually. This study proposes a real-time water level detection system based on a computer vision algorithm. In the proposed system, we use color object tracking technique with a bar indicator as a reference’s level. We set three bar indicators to determine the status of the water level, namely NORMAL, ALERT and DANGER. A camera was installed across the bar level indicators to capture bar indicator and monitoring the water level. In the simulation, the monitoring system was installed in 5-100 lux lighting conditions. For experimental purposes, we set various distances of the camera, which is set of 40-80 centimeters and the camera angle is set of 30-60 degrees. The experiment results showed that this system has an accuracy of 94% at camera distance is in range 50-80 centimeters and camera angle is 60o. Based on these results, it can be concluded that this proposed system can determine the water level well in varying lighting conditions.

Integrating YOLOv7 with FixMatch for Enhancing Vehicle Detection Performance in Mixed Traffic Environments Zaita, Yandri; Saddami, Khairun; Nasaruddin, Nasaruddin
JURNAL NASIONAL TEKNIK ELEKTRO Vol 13, No 3: November 2024
Publisher : Jurusan Teknik Elektro Universitas Andalas

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25077/jnte.v13n3.1250.2024

A major challenge in the development of object detection technology is the significant reliance on large labeled datasets, which requires substantial time and memory for manual annotation—especially in complex, mixed traffic environments with varied vehicle types, congestion levels, and unpredictable motion patterns. This study addresses this issue by integrating the semi-supervised learning technique, FixMatch, into the YOLOv7 object detection model, utilizing 4000 transportation-related datasets. The FixMatch technique enables the model to detect unlabeled objects effectively through strong and weak augmentation methods. In this study, the detected objects in the mixed traffic environment include public transportation, pedicabs, cars, motorcycles, and trucks. This study achieved an impressive 97.5% detection accuracy by leveraging unlabeled data, demonstrating the model's efficiency and effectiveness in identifying vehicles under diverse traffic conditions. Consequently, integrating the FixMatch method into YOLOv7 provides a practical and efficient solution for object detection in situations where collecting labeled data is challenging, such as in dynamic and highly variable traffic environments.

Title

Found 13 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 13 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 13 Documents
Search