JOIV : International Journal on Informatics Visualization
Not a member yet
786 research outputs found
Sort by
A Robust License Plate Detection System Using Smart Device
Get PDFThe license plate recognition (LPR) system is widely employed in various applications. However, most research studies have used a fixed camera rather than a moving one. This is because the location of the vehicle plate is nearly static and easily estimated, making the use of a static camera simple for locating and detecting the scanned license plate. Images obtained with a moving camera are highly complex due to frequent background changes. Additionally, a challenge with car plates in Malaysia is their non-standardized nature. Car owners are permitted to use any font type for their license plate number, rendering existing license plate recognition systems from other countries incapable of effectively detecting license plates on Malaysian car plates. A traditional LPR system typically requires a high-quality camera and a powerful computer for costly and bulky processing. Nowadays, many smartphones come equipped with powerful processors and cameras. Android smartphones include various libraries for modifying hardware configurations such as the camera. This paper presents a robust method for detecting Malaysia's license plate number using a convolutional neural network (CNN). The CNN model from the pre-training process is imported to the Android device and tested in real-time in an on-road driving environment, resulting in an average recognition rate of 89.37%. A comprehensive Character Recognition Analysis is also presented to demonstrate the accuracy of each character. However, there is still room for improvement in recognizing the character Q
A Comparative Study of Image Retrieval Algorithm in Medical Imaging
Get PDFIn recent times, digital environments have become more complex, and the need for secure, efficient, and reliable identification systems is growing in demand. Consequently, image retrieval has emerged as a critical area focusing on artificial intelligence and machine learning applications. Medical image retrieval has become increasingly crucial in today's healthcare field, as it involves accurate diagnostics, treatment planning, and advanced medical research. As the quantity of medical imaging data grows rapidly, the ability to efficiently and accurately retrieve relevant images from extensive datasets becomes critical. Advanced retrieval systems, such as content-based image retrieval, are imperative for managing complex data, ensuring that healthcare professionals can access the most relevant information to improve patient outcomes and advance medical knowledge. This paper compares three algorithms: Scale Invariant Feature Transform, Speeded Robust Features, and Convolutional Neural Networks in the context of two medical image datasets, ImageCLEF and Unifesp. The findings highlight the trade-offs between precision and recall for each algorithm, providing invaluable insights into selecting the most suitable algorithm for specific tasks. The study evaluates the algorithms based on precision and recall, two critical performance metrics in image retrieval
Analyzing Rupiah-USD Exchange Rate Dynamics: A Study with ARCH and GARCH Models
Get PDFThe study aims to analyze the volatility of the Rupiah-USD exchange rate and predict future fluctuations using the Autoregressive Conditional Heteroskedasticity (ARCH) and Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models. The exchange rate data, spanning from January 2010 to December 2023, is sourced from Bank Indonesia (BI) and adheres to the Jakarta Interbank Spot Dollar Rate (JISDOR) regulations, focusing solely on business days. ARCH and GARCH models are widely applied in financial time series analysis because they capture and forecast time-varying volatility. This study analyzes historical exchange rate data to evaluate the persistence of volatility and detect any structural breaks that could impact future exchange rate behavior. The findings reveal that both models effectively capture the volatility of the Rupiah-USD exchange rate, but the GARCH (1,1) model demonstrates superior forecasting accuracy. This model's ability to account for long-term volatility clustering makes it particularly useful for predicting exchange rate dynamics. The research contributes to a deeper understanding of the factors driving exchange rate fluctuations, offering valuable insights for policymakers, investors, and businesses. These insights can help stakeholders manage exchange rate risks more effectively within Indonesia's open economy, where global financial conditions and external shocks significantly shape currency movements. The study emphasizes the importance of using advanced econometric models for accurate volatility predictions and informed decision-making
Developing and Comparing Machine Learning Algorithms for Music Recommendation
Get PDFThe increasing prevalence of song skipping in music streaming applications negatively impacts user satisfaction and subscription retention. Dissatisfaction often arises when users encounter songs they actively dislike, highlighting a gap between user expectations and the value offered by these services. To address this, music recommendation algorithms were researched and developed. Initially, data collection is started. Data collection is through the Spotify application programming interface. This initiation step sets the stage for subsequent exploratory data analysis. Exploratory data analysis examined the collected data to plot a bar chart for total songs released over the years, plot a bar chart for the popularity of songs based on the year it is released, visualize word cloud on frequently mentioned music genres, chart count plot for explicit songs, and chart count plot for song modalities. Data preprocessing involved cleaning the data, handling missing values, and checking for null values to prepare the application of machine learning algorithms. Four machine learning algorithms were applied, k-means, mini-batch k-means, Gaussian mixture, and density-based spatial clustering of applications with noise (DBSCAN), to analyze musical features like rhythm, tempo, and other relevant music attributes. The results showed that the k-means outperforms all other algorithms evaluated regarding recommendation quality, as measured by the Calinski-Harabasz score. Based on the evaluation, the best machine learning will then be applied to identify similarities between songs and be used to generate music recommendation results
Development of Smart Simulator for Electronic Fuel Injection (EFI) fuel system based on Quick Response Code (QR Code) for Learning Media
Get PDFThis research aims to develop learning media in the form of an intelligent fuel system simulator, namely electronic fuel injection (EFI) on motorbikes, which focuses on developing the addition of Quick Response (QR) codes as additional information from the simulator, which is connected to video on the YouTube platform. With the help of the QR code, it is scanned using a smartphone so that the QR code can be connected directly to YouTube, providing additional information about the Smart simulator EFI system on motorbikes. This research was carried out by applying the Research and Development method and following the Plomp model, which consists of the following stages: (1) preliminary investigation, (2) designing and making a prototype, and (3) assessment. The existing simulator was developed by adding a QR code, and the QR code will be connected to videos that have been uploaded on the YouTube platform. QR codes are created using the online QR code generator platform. Assessment of the smart simulator is carried out through a questionnaire filled out by media experts and subject experts, as well as through observation sheets during smart simulator testing. The research results are a smart simulator product for EFI fuel on motorbikes equipped with a QR code. Evaluation by media and material experts shows that the smart simulator is declared valid. Meanwhile, the results of observations during product testing show that the smart simulator can describe the characteristics of the EFI fuel system on a motorbike according to the actual situation
A Thorough Review of Vehicle Detection and Distance Estimation Using Deep Learning in Autonomous Cars
Get PDFAutonomous vehicle technologies are rapidly advancing, and one key factor contributing to this progress is the enhanced precision in vehicle detection and distance calculation. Deep Learning Networks (DLNs) have emerged as powerful tools to address this challenge, offering remarkable capabilities in accurately detecting and estimating vehicle positions. This study comprehensively reviews DLN applications for vehicle detection and distance estimation. It examines prominent DLN models such as YOLO, R-CNN, and SSD, evaluating their performance on widely used datasets such as KITTI, PASCAL VOC, and COCO. Analysis results indicate that YOLOv5, developed by Farid et al. achieves the highest accuracy level with a mAP (mean Average Precision) of 99.92%. Yang et al. showcased that YOLOv5 performs exceptionally in detection and distance estimation tasks, with a mAP of 96.4% and a low mean relative error (MRE) of 10.81% for distance estimation. These achievements highlight the potential of DLNs to enhance the accuracy and reliability of vehicle detection systems in autonomous vehicles. The study also emphasizes the importance of backbone architectures like DarkNet 53 and ResNet in determining model efficiency. The choice of the appropriate model depends on the specific task requirements, with some models prioritizing real-time detection and others prioritizing accuracy. In conclusion, developing DLN-based methods is crucial in advancing autonomous vehicle technology. Research and development remain crucial in ensuring road safety and efficiency as autonomous vehicles become more common in transportation systems
Wireless Volume Corrector for Natural Gas Flow Metering Using ESP32 Microcontroller and Open-Source Web Server
Get PDFThe measurement of gas flow in energy transactions from an energy-producing company to an energy user is essential to evaluate, considering that measurement errors can accumulate, and losses can occur, which will be ongoing. The gas measurement process begins with measuring pressure and temperature parameters and the gas flow volume transacted. Then, the measured gas volume will be converted into a standard volume as the basis for gas buying and selling transactions (custody). This article conveys that a wireless volume corrector has been designed using the ESP32 microcontroller with Wi-Fi communication that can be integrated with the internet to support automatic meter reading systems. The measured data is collected in a database server using XAMPP, an open-system web server. Data stored on the database server is displayed via a web browser with a display of the last 5 data entered the web server. The web browser display is refreshed every second so that the display on the web browser is a live or online display. This research resulted in temperature, pressure, actual flow, and volume standard measurements with an error below 0.1%, which met the metrology requirements, instilling confidence in the system's reliability. This research proves that realizing a volume corrector with IoT technology can be done cheaply
Skin Lesion Classification: A Deep Learning Approach with Local Interpretable Model-Agnostic Explanations (LIME) for Explainable Artificial Intelligence (XAI)
Get PDFThe classification of skin cancer is crucial as the chance of survival increases significantly with timely and accurate treatment. Convolution Neural Networks (CNNs) have proven effective in classifying skin cancer. However, CNN models are often regarded as "black boxes”, due to the lack of transparency in the decision-making. Therefore, explainable artificial intelligence (XAI) has emerged as a tool for understanding AI decisions. This study employed a CNN model, VGG16, to classify five skin lesion classes. The hyperparameters were adjusted to optimize its classification performance. The best hyperparameter settings were 50 epochs, a 0.1 dropout rate, and the Adam optimizer with a 0.001 learning rate. The VGG16 model demonstrated satisfactory classification performance. The Local Interpretable Model-Agnostic Explanations (LIME) method was implemented as the XAI tool to justify the predictions made by VGG16. The LIME explanation revealed that the correct predictions made by VGG16 were owing to its truthful extraction of the cancer or lesion area, especially for the “vascular lesion” class. Meanwhile, inaccurate classifications were attributed to VGG16 extraction of the background and insignificant parts of the skin as core features. In conclusion, The LIME model allowed visual inspection of the features selected by VGG16, paving the way for improving the CNN model for better feature extraction and classification of skin lesions, offering a promising direction for future research.
Multimedia-Assisted Elementary School Learning Materials Innovation Using STEAM Learning Approach
Get PDFPrinted materials dominate the teaching materials used in elementary schools in Indonesia. These conditions are backward with the development of the Industrial Revolution 4.0, so it is urgent to research the effectiveness of multimedia-based teaching material using Macromedia Flash Eight application with the learning approach of Science, Technology, Engineering, Art, and Math (STEAM) in Elementary Schools. The research method used is research and development (R&D) with the ADDIE model consisting of Analysis, Design, Development, Implementation, and Evaluation. The data collection instrument uses pre-test and post-test sheets, and an evaluation lift is open to the use of teaching materials for teachers and students. The respondents were eight teachers and 84 elementary school students. The effectiveness of the teaching material is analyzed in two ways: (1) the analysis of learning outcomes (pre-test and post-test) using the N-Gain Test, and (2) the evaluation of the product is based on the results of open-haul analysis, advice, and recommendations. This research and development results show that the pre-test average of 47,14 increased to 84.58 (post-test), while the test score of N-Gain was 71.75%, with the category quite effective. Thus, it can be concluded that Macromedia Flash 8-based teaching materials with STEAM approach to integrated thematic learning in elementary schools are practical and capable of improving student learning outcomes
Implementing K-Nearest Neighbors (k-NN) Algorithm and Backward Elimination on Cardiotocography Datasets
Get PDFHaving a healthy baby is a dream for mothers. Unfortunately, high maternal and fetal mortality has become a vital problem that requires early risk detection for pregnant women. A cardiotocograph examination is necessary to maintain maternal and fetal health. One method that can solve this problem is classification. This research analyzes the optimal use of k values and distance measurements in the k-NN method. This research expects to become the primary reference for other studies examining the same dataset or developing k-NN. A selection feature is needed to optimize the classification method, particularly for improving accuracy results. This study used the cardiotocography dataset from cardiotocograph examinations related to fetal conditions. The cardiotocography dataset consisted of 2,126 records with 22 features and variables. It also had three classification classes, normal, suspect, and pathological, from the Universal Child Immunization Machine Learning Repository website. It employed the K-Nearest Neighbor (k-NN) method and the backward elimination feature with ordinary least squares regression. The test in this research applied the scenarios of three distance calculations, i.e., Euclidean distance, Manhattan distance, and Minkowski distance, as well as four variations of k values. Evaluation of each scenario indicated the accuracy of the confusion matrix and execution time. This study compared K-Nearest Neighbor (k-NN) and Backward Elimination methods with K-nearest neighbor (k-NN) without selection features. The best accuracy of the Backward Elimination and K-Nearest Neighbor (K-NN) methods was 91%, as was the K-Nearest Neighbor (k-NN) method without selection features. Both had similar k values (k = 3) and Manhattan distance. The backward elimination method reduced the number of features from 22 to 14. Meanwhile, the execution times of the Backward Elimination and K-Nearest Neighbor (k-NN) methods got better results as each distance averaged 26.54, 19.23, and 68.09 seconds. K-Nearest Neighbor (k-NN) execution times without selection features were 26.83, 19.39, and 68.84, respectively. In conclusion, backward elimination did not increase accuracy because it yielded the same accuracy. However, backward elimination and K-nearest Neighbor (k-NN) produced faster results, with differences of 29%, 16%, and 75%, respectively