Claim Missing Document
Check
Articles

Found 33 Documents
Search

Comparison of two deep learning methods for detecting fire hotspots Dewi Putrie Lestari; Rifki Kosasih
International Journal of Electrical and Computer Engineering (IJECE) Vol 12, No 3: June 2022
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijece.v12i3.pp3118-3128

Abstract

Every high-rise building must meet construction requirements, i.e. it must have good safety to prevent unexpected events such as fire incident. To avoid the occurrence of a bigger fire, surveillance using closed circuit television (CCTV) videos is necessary. However, it is impossible for security forces to monitor for a full day. One of the methods that can be used to help security forces is deep learning method. In this study, we use two deep learning methods to detect fire hotspots, i.e. you only look once (YOLO) method and faster region-based convolutional neural network (faster R-CNN) method. The first stage, we collected 100 image data (70 training data and 30 test data). The next stage is model training which aims to make the model can recognize fire. Later, we calculate precision, recall, accuracy, and F1 score to measure performance of model. If the F1 score is close to 1, then the balance is optimal. In our experiment results, we found that YOLO has a precision is 100%, recall is 54.54%, accuracy is 66.67%, and F1 score is 0.70583667. While faster R-CNN has a precision is 87.5%, recall is 95.45%, accuracy is 86.67%, and F1 score is 0.913022.
Detection and Classification of Vehicles on the Bekasi Toll Road Using the Gaussian Mixture Models Method and Morphological Operations Rifki Kosasih; Hidayat Taufik Akbar
Telematika Vol 15, No 1: February (2022)
Publisher : Universitas Amikom Purwokerto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35671/telematika.v15i1.1222

Abstract

Traffic surveillance was initially carried out directly using CCTV, but this kind of surveillance was not possible for a full day by the security forces. In addition, with the increasing growth of vehicles in Indonesia, a method is needed that can be used to assist security forces in monitoring traffic such as detecting and automatically counting the number of vehicles. Therefore, in our research, we propose a method that can detect vehicles, and count the number of vehicles from video recordings on the Bintara Bekasi toll road using background substraction methods such as gaussian mixture models and morphological operations. The results showed that the vehicle detection accuracy rate was 86.3636%, the precision was 89.0625%, and the recall was 96.6101%. In this study, vehicle classification was also carried out based on the detection results into two types of vehicles, namely cars and trucks. From the results of the research, the classification accuracy rate was obtained at 85.9649%.
Kombinasi Metode ISOMAP Dan KNN Pada Image Processing Untuk Pengenalan Wajah Rifki Kosasih
CESS (Journal of Computer Engineering, System and Science) Vol 5, No 2 (2020): JULI 2020
Publisher : Universitas Negeri Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (651.855 KB) | DOI: 10.24114/cess.v5i2.18982

Abstract

Wajah manusia memiliki ciri khusus yang dapat membedakan dengan orang lainnya sehingga pengenalan wajah sangat penting dilakukan untuk mengenali seseorang. Ciri khusus pada wajah ini disebut juga dengan fitur. Pada penelitian ini, untuk mendapatkan fitur dilakukan pengenalan pola citra wajah dengan menggunakan metode isomap. Metode isomap merupakan salah satu metode dari manifold learning yang menghasilkan fitur-fitur dengan cara mereduksi dimensi. Citra wajah yang digunakan dalam penelitian ini terdiri dari 6 orang dengan tiap orang memiliki 4 citra wajah dengan ekspresi yang berbeda-beda. Data citra ini dibagi menjadi dua bagian yaitu data latih dan data uji. Selanjutnya data citra tersebut diubah menjadi vektor. Metode isomap digunakan untuk mentransformasikan vektor tersebut menjadi vektor yang mengandung fitur wajah. Setelah fitur wajah diperoleh, selanjutnya dilakukan pengujian pada data uji dengan menggunakan algoritma K Nearest Neighbor (KNN).  Algoritma K Nearest Neighbor digunakan untuk pengklasifikasian dengan cara mencari K data latih yang terdekat dengan data uji. Dari hasil klasifikasi diperoleh bahwa tingkat akurasi sebesar 83,33%.
Analisis Sentimen Produk Permainan menggunakan Metode TF-IDF dan Algoritma K-Nearest Neighbor Rifki Kosasih; Anggi Alberto
InfoTekJar : Jurnal Nasional Informatika dan Teknologi Jaringan Vol 6, No 1 (2021): InfoTekJar September
Publisher : Universitas Islam Sumatera Utara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30743/infotekjar.v6i1.3893

Abstract

Pada situs belanja online, terdapat kolom komentar atau rating dari pembeli yang telah melakukan transaksi pada produk tersebut. Dengan adanya fitur penilaian produk berdasarkan rating tersebut, pihak pembeli dapat mengetahui seberapa baik atau buruknya produk tersebut. Akan tetapi muncul permasalahan dimana ada beberapa pembeli memberikan komentar negatif dengan rating sebesar lima bintang ataupun sebaliknya, hal tersebut menyebabkan fitur penilaian produk berdasarkan rating menjadi kurang baik. Oleh karena itu untuk dapat mengetahui kualitas produk tersebut dilakukan analisis sentimen dengan metode TF-IDF dan K-Nearest Neighbor (KNN) berdasarkan ulasan dari pembeli. Data yang dikumpulkan adalah 1000 ulasan yang dibagi menjadi 700 data latih dan 300 data uji. Tahapan selanjutnya dilakukan teks preprocessing seperti case folding (mengubah huruf besar menjadi kecil), tokenizing (pemisahan kalimat menjadi kata tunggal), stopword (menghilangkan kata sambung hasil tokenizing yang tidak ada hubungannya dalam analisis sentimen),  stemming (mengubah kata ke bentuk kata dasar) dan pembobotan kata dengan TF-IDF. Tahapan terakhir adalah melakukan klasifikasi dengan menggunakan metode K Nearest Neighbor (K-NN). Berdasarkan hasil klasifikasi diperoleh tingkat akurasi sebesar 79,3333%.
Pengenalan Wajah Menggunakan PCA dengan Memperhatikan Jumlah Data Latih dan Vektor Eigen Rifki Kosasih
Jurnal Informatika Universitas Pamulang Vol 6, No 1 (2021): JURNAL INFORMATIKA UNIVERSITAS PAMULANG
Publisher : Teknik Informatika Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/informatika.v6i1.7261

Abstract

To find out if an employee is present, attendance is usually used. Attendance can be done in several ways, one of which is by filling in the attendance list that has been provided (manual attendance). However, this method is less effective because there is a possibility that employees who are not present will entrust attendance to employees who are present. Therefore, other ways are needed so that this does not happen. In this study, attendance was carried out using facial recognition. Face recognition is one of the fields used to recognize someone. A person's face usually has special characteristics that are easily recognized by people. These special characteristics are also called features. In this study, these features can be searched using the Principle Component Analysis (PCA) method. The PCA method is one of the methods used to produce features by reducing dimensions using eigenvectors from facial images (eigenface). The facial image used in this study consisted of 40 people with each person having 10 facial images with various expressions. Image data is divided into two parts, namely training data and test data. In this study, it is proposed to pay attention to the amount of training data and the number of eigenvectors used to get the best level of accuracy. From the research results, the highest level of accuracy occurs when the training data for each person is 7 and the test data for each person is 3 with an accuracy rate of 96.67%.
Sentiment analysis of game product on shopee using the TF-IDF method and naive bayes classifier Rifki Kosasih; Anggi Alberto
ILKOM Jurnal Ilmiah Vol 13, No 2 (2021)
Publisher : Teknik Informatika Fakultas Ilmu Komputer Univeristas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v13i2.721.101-109

Abstract

In every product sold on the E-commerce platform, there is a review column from consumers who have made transactions on the products. These reviews are in the form of comments and ratings (stars from one to five) written and given by consumers based on their assessment of the products purchased. With the product evaluation feature based on the rating, the consumer can find out how good or bad the quality of the product is. However, a problem arises when some consumers give negative comments with five stars or vice versa. This causes the product assessment feature based on the rating to be less good so that it does not represent the real value. Therefore, to determine the quality of the product, sentiment analysis was carried out using the TF-IDF method and the Naive Bayes Classifier based on reviews from buyers. The data collected is 1000 reviews which are divided into 700 training data and 300 test data. The next stage is the preprocessing text such as case folding (converting uppercase letters to lowercase), tokenizing (separating sentences into single words), stopwords (removing tokenizing conjunctions that have nothing to do with sentiment analysis), stemming (changing words into basic word forms), and word weighting with TF-IDF. The last step is to classify. Based on the classification results obtained an accuracy rate of 80.2223%.
Pengenalan Wajah dengan Menggunakan Metode Local Binary Patterns Histograms (LBPH) Rifki Kosasih; Christian Daomara
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 5, No 4 (2021): Oktober 2021
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v5i4.3171

Abstract

Rapid technological developments in the world have had a major impact on various fields, i.e. administration and data collection. One of the data collection systems that is often used is attendance. The attendance system initially only used an attendance sheet that was filled out manually, but there were shortcomings i.e. the possibility of forging signatures during attendance. Therefore, other methods are needed to overcome these problems. In this study, we propose to use facial features for attendance. The data used are 750 facial images consisting of 5 people with each person having 150 facial images with various expressions. The data is divided into two, namely 500 images as training data and 250 images as test data. The next stage is to find a facial features, we propose to use the Local Binary Patterns Histograms (LBPH) method. LBPH is a combination of the Local Binary Patterns (LBP) method with Histograms of Oriented Gradients (HOG). After that, we perform face recognition based on the features that have been obtained. Based on the research results obtained an accuracy rate of 86%
Pendeteksian Kendaraan dengan Menggunakan Metode Running Average Background Substraction dan Morfologi Citra Rifki Kosasih; Muhammad Arfiansyah
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 4, No 4 (2020): Oktober 2020
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v4i4.2315

Abstract

Traffic conditions on the highway at this time has started to be crowded. To find out if there is traffic jam or can not be seen by counting the number of vehicles passing through the area. However, it is impossible for security forces to count the number of vehicles manually. This requires a method that can be used and is calculated from the number of vehicles. In this study, the running average background substraction method and morphological operations were used to detect vehicles and use the center of the object (centroid) to calculate the number of vehicles. The sample used is a traffic video in the Bekasi area. From the research results, there were 37 vehicles detected in real conditions and stated as vehicles in the application and there were 7 vehicles detected in real conditions but not stated in the application. The next stage is an evaluation by calculating the value of precision, recall and accuracy. In this study, the precision value obtained was 84.09%, the recall value was 94.87% and the accuracy rate was 80.43%.
Klasifikasi Tingkat Kematangan Pisang Berdasarkan Ekstraksi Fitur Tekstur dan Algoritme KNN Rifki Kosasih
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 10 No 4: November 2021
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1138.146 KB) | DOI: 10.22146/jnteti.v10i4.462

Abstract

Bananas are fruits that are rich in vitamins, minerals, and carbohydrates. Banana trees are often cultivated as they have many benefits. In growing banana trees, it is necessary to consider the ripeness level of bananas since it can determine the quality of bananas when harvested. The ripeness level of bananas is related to marketing reach. If the marketing reach is far, the banana should be harvested when it is still raw. Therefore, a system that can classify bananas’ ripeness levels is needed. In this study, 45 banana images were collected, with a composition of 30 images as training data and 15 images as test data. Afterwards, the texture feature extraction method was utilized to determine the parameters affecting the ripeness level of bananas. The texture feature extraction used was based on a histogram that generated several parameters i.e., average intensity, skewness, energy descriptor, and smoothness in the image. In the subsequent stage, the classification based on the features obtained using KNN algorithm was conducted. Based on the results, it was found that the classification accuracy rate was 88.89%.
Implementation of Random Forest on Face Recognition Using Isomap Features Rifki Kosasih; Achmad Fahrurozi; Desti Riminarsih
CESS (Journal of Computer Engineering, System and Science) Vol 7, No 2 (2022): July 2022
Publisher : Universitas Negeri Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24114/cess.v7i2.34498

Abstract

Sistem pengenalan wajah merupakan salah satu bidang yang digunakan untuk mengenali wajah seseorang. Dalam penelitian ini, data yang dikumpulkan merupakan data citra wajah yang terdiri dari 24 citra dengan komposisi 6 orang dan tiap orang memiliki 4 citra dengan berbagai ekspresi. Untuk mengenali wajah tersebut, dilakukan ekstraksi fitur wajah terlebih dahulu menggunakan metode isomap. Isomap merupakan metode reduksi dimensi yang dapat mereduksi dari dimensi tinggi menjadi fitur-fitur yang berdimensi rendah. Berdasarkan hasil ekstraksi diperoleh 4 fitur yang digunakan untuk mengklasifikasikan wajah. Untuk mengklasifikasikan wajah, digunakan algoritma random forest. Berdasarkan hasil penelitian diperoleh tingkat akurasi hasil klasifikasi sebesar 87,5%, nilai weighted average precision sebesar 81,25% dan nilai weighted average recall sebesar 87,5%.