Indonesian Journal of Physics

Multiclass Classification of Covid-19 CT Scan Images With VGG-16 Architecture Using Transfer Learning System

2024-12-10T01:38:13+00:00

COVID-19 is a respiratory disease caused by the coronavirus. The most common test technique used today for COVID-19 diagnosis is real-time reverse transcription-polymerase chain reaction (RT-PCR). However, compared to RT-PCR, radiological imaging such as X-rays and computer tomography (CT) may be a more precise, useful, and faster technology for COVID-19 classification. X-rays are more accessible because they are widely available in all hospitals in the world and are cheaper than CT scans, but the classification of COVID-19 using CT scan images is more sensitive than X-rays. Therefore, CT scan images can be used for the early detection of COVID-19 patients. One of them is using the deep learning method. In this study, a CNN algorithm with a VGG-16 architecture will be selected to classify COVID-19, intermediate, and non-COVID CT scan images using 2481 image datasets. First, pre-processing is done by resizing the image, converting the image channel into RGB, and dividing the dataset into a training dataset and a testing dataset. Then, the convolution process is continued by utilizing the pre-trained VGG-16 model from ImageNet. The results of testing the data with 97% accuracy were obtained. It is concluded that the model used to classify COVID-19, intermediate, and non-COVID CT scan images is effective and produces good results.

Mechanical Traction Test of Pineapple Leaf Fiber-Epoxy Composite

2024-12-09T11:16:59+00:00

Natural fiber-based composites are widely developed to replace materials in various industries, such as furniture and sound absorbers. One of the composite properties that is often considered for its use is mechanical property. In this study, the mechanical properties of pineapple leaf fiber-epoxy composites were studied using traction tests. The composite was made by mixing of small pieces of pineapple leaf fiber with epoxy resin, then the mixture was placed in a mold and pressed with a certain pressure at 100 °C for 2 hours. Traction tests were carried out on the pineapple leaf fiber-epoxy composites with two different epoxy concentrations (11 and 17 % wt), and different manufacturing pressure (0.22; 0.65; 1.09 MPa). The traction test results show that the maximum stress (ultimate tensile strength/UTS), elongation at break (EAB) and modulus of elasticity (E) increase with increasing epoxy content. Increasing the manufacturing pressure increases modulus of elasticity. However, the maximum UTS and EAB were shown by the composite with a manufacturing pressure of 0.65 Pa, not at the higher pressure (1.09 Pa).

Object’s Movement Simulation with Air Drag: Aerodynamics Wall and Knuckle’s Effect

2024-12-10T02:01:55+00:00

The drag force of air and objects can be analyzed using the Stokes or Quadratic, also known as the Newtonian method. In this research, a Newtonian model was created numerically in Python using a 4^th order Runge-Kutta integrator. The integrator will solve the acceleration function experienced by the object when given air resistance into a position function. The object’s movement influenced by variation of drag’s coefficient will provide variations in the location of the aerodynamic wall, the condition when the object moves vertically downwards. The movement of an object that is influenced by air resistance under certain conditions will also experience a lifting force, due to the rotational orientation angle of the object. The movement of objects influenced by these two forces can be modeled as knuckle movement. The results obtained from the numerical solution of the two object conditions are then made into a simple application in the form of a GUI so that users can easily operate the object simulation. This research will be a solution for students to understand the motion of objects influenced by the drag forces that occur when taking classical mechanics courses.

A Review on Exploring the resonant vibration of thin plates: Reconstruction of Chladni patterns and determination of resonant wave numbers Using Research Based Learning

2024-12-10T01:59:16+00:00

Belajar adalah proses perolehan pengetahuan atau keterampilan melalui studi atau pengalaman. Salah satu metode

pembelajaran untuk meningkatkan pemahaman siswa adalah Research-based Learning (RBL), sebuah metodologi yang bertujuan

menanamkan berfikir kritis, kreativitas, dan keterampilan memecahkan masalah. Pada penelitian ini, metode RBL digunakan untuk

meningkatkan pemahaman dalam percobaan Chladni dengan mengamati fenomena resonansi ketika pelat diberi getaran dengan

frekuensi tertentu serta menentukan besar frekuensi yang menghasilkan pola-pola Chladni. Pada metode ini, siswa akan bekerja

dalam kelompok yang terdiri dari 3 orang. Ada lima langkah yang akan dilakukan pada RBL ini: (i) merumuskan masalah dan

tujuan, membuat hipotesis, (ii) melakukan eksperimen, mengoleksi data hasil eksperimen (iii) mencari informasi mengenai

eksperimen, memproses data (iii) menganalisis data, menjawab rumusan masalah (iv) membuat kesimpulan, (v) mempresentasikan

hasilnya kepada dosen/penguji. Pada akhir eksperimen ini siswa mampu membuktikan hukum Chladni.

Application of YOLOv5s Algorithm for Real-Time Object Detection in Mobile Robot for Volcano Monitoring System

2024-12-10T02:00:59+00:00

Indonesia, a country with 172 volcanoes and second after Japan for the most eruption events, should monitor and predict the volcano eruption to prevent the effect of this natural disaster. Therefore, we have developed a 4-wheeled mobile robot equipped with monitoring sensors and a Logitech camera for this purpose. The robot should have the ability to detect objects in this extreme environment to avoid collision while moving and monitoring the volcano’s physical parameters. It has been designed a deep machine learning of YOLOv5s algorithm for two objects mostly found at volcanoes such as trees and stones. After the training steps (object identification; dataset downloading (Google Chrome Extension and Open Images v6); image labeling (LabeImg); augmentation process (blur and rotation)) had been carried out, the images of the object then trained in three model variation which resulted in: mAP_0.5 = 51.9%, mAP_0.5:0.95 = 28.6%, 58% of precision and 50% recall with 12 minutes and 33 seconds of training time for the first model (batch=16 and epochs=100); mAP_0.5 = 59.7%, mAP_0.5:0.95 = 36.3%, 74% of precision and 54% recall with 36 minutes and 4 seconds of training time for the second model (batch=16 and epochs=300); mAP_0.5 = 59.9%, mAP_0.5:0.95 = 37.6%, 80% of precision and 55% recall with one hour and 25 seconds of training time for the last one (batch=16 and epochs=500) as the best model of these variations. Furthermore, these results were displayed for all test images for the best model.