• Geomechanics and Engineering
• /
• v.38 no.1
• /
• pp.29-44
• /
• 2024

This research investigates the effectiveness of data augmentation techniques in the automated analysis of B-scan images from ground-penetrating radar (GPR) using deep learning. In spite of the growing interest in automating GPR data analysis and advancements in deep learning for image classification and object detection, many deep learning-based GPR data analysis studies have been limited by the availability of large, diverse GPR datasets. Data augmentation techniques are widely used in deep learning to improve model performance. In this study, we applied four data augmentation techniques (geometric transformation, color-space transformation, noise injection, and applying kernel filter) to the GPR datasets obtained from a testbed. A deep learning model for GPR data analysis was developed using three models (Faster R-CNN ResNet, SSD ResNet, and EfficientDet) based on transfer learning. It was found that data augmentation significantly enhances model performance across all cases, with the mAP and AR for the Faster R-CNN ResNet model increasing by approximately 4%, achieving a maximum mAP (Intersection over Union = 0.5:1.0) of 87.5% and maximum AR of 90.5%. These results highlight the importance of data augmentation in improving the robustness and accuracy of deep learning models for GPR B-scan analysis. The enhanced detection capabilities achieved through these techniques contribute to more reliable subsurface investigations in geotechnical engineering.

• Journal of Multimedia Information System
• /
• v.6 no.4
• /
• pp.165-172
• /
• 2019

Machine-learning techniques have been actively employed to information security in recent years. Traditional rule-based security solutions are vulnerable to advanced attacks due to unpredictable behaviors and unknown vulnerabilities. By employing ML techniques, we are able to develop intrusion detection systems (IDS) based on anomaly detection instead of misuse detection. Moreover, threshold issues in anomaly detection can also be resolved through machine-learning. There are very few datasets for network intrusion detection compared to datasets for malicious code. KDD CUP 99 (KDD) is the most widely used dataset for the evaluation of IDS. Numerous studies on ML-based IDS have been using KDD or the upgraded versions of KDD. In this work, we develop an IDS model using CSE-CIC-IDS 2018, a dataset containing the most up-to-date common network attacks. We employ deep-learning techniques and develop a convolutional neural network (CNN) model for CSE-CIC-IDS 2018. We then evaluate its performance comparing with a recurrent neural network (RNN) model. Our experimental results show that the performance of our CNN model is higher than that of the RNN model when applied to CSE-CIC-IDS 2018 dataset. Furthermore, we suggest a way of improving the performance of our model.

• IEIE Transactions on Smart Processing and Computing
• /
• v.4 no.2
• /
• pp.110-114
• /
• 2015

This paper reports a novel object-detection technique in video sequences. The proposed algorithm consists of detection of objects in illuminated and occluded videos by using object features and a neural network technique. It consists of two functional modules: region-based object feature extraction and continuous detection of objects in video sequences with region features. This scheme is proposed as an enhancement of the Lowe's scale-invariant feature transform (SIFT) object detection method. This technique solved the high computation time problem of feature generation in the SIFT method. The improvement is achieved by region-based feature classification in the objects to be detected; optimal neural network-based feature reduction is presented in order to reduce the object region feature dataset with winner pixel estimation between the video frames of the video sequence. Simulation results show that the proposed scheme achieves better overall performance than other object detection techniques, and region-based feature detection is faster in comparison to other recent techniques.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.5
• /
• pp.630-636
• /
• 2020

This study reviewed camouflage techniques to reduce the visual detect-ability of aircraft. Visual camouflage can be defined as the process of making objects less visible. Aircraft visual camouflage delays detection of the aircraft position, speed, and flight direction. Multi-tone and counter-shaded schemes are generally adopted as camouflage patterns for close-air-support aircraft and air-superiority aircraft, respectively. Another study showed that the monotone scheme is also efficient when the hue and brightness of the camouflage color are controlled correctly. Active camouflage techniques for aircraft have been studied to increase the camouflage effectiveness. In particular, counter-illumination techniques using electroluminescence devices can minimize the difference in brightness between the aircraft and sky background. Active camouflage techniques are expected to enhance the survivability of low-altitude UAVs, which are vulnerable to visual detection.

• Journal of KIISE:Information Networking
• /
• v.31 no.2
• /
• pp.171-178
• /
• 2004

Port scanning detection systems should rather satisfy a certain level of the requirement for system performance like a low rate of “False Positive” and “False Negative”, and requirement for convenience for users to be easy to manage the system security with detection systems. However, public domain Real Time Scan Detection Systems have high rate of false detection and have difficulty in detecting various scanning techniques. In addition, as current real time scan detection systems are based on command interface, the systems are poor at user interface and thus it is difficult to apply them to the system security management. Hence, we propose TkRTSD(Tcl/Tk Real Time Scan Detection System) that is able to detect various scan attacks based on port scanning techniques by applying a set of new filter rules, and minimize the rate of False Positive by applying proposed ABP-Rules derived from attacker's behavioral patterns. Also a GUI environment for TkRTSD is implemented by using Tcl/Tk for user's convenience of managing network security.

• KIISE Transactions on Computing Practices
• /
• v.23 no.2
• /
• pp.128-133
• /
• 2017

In recent years, the number of systems for the analysis of large volumes of data is increasing. Hadoop, a representative big data system, stores and processes the large data in the distributed environment of multiple servers, where system-resource management is very important. The authors attempted to detect anomalies from the rapid changing of the log data that are collected from the multiple servers using simple but efficient anomaly-detection techniques. Accordingly, an Apache Hive storage architecture was designed to store the log data that were collected from the multiple servers in the Hadoop ecosystem. Also, three anomaly-detection techniques were designed based on the moving-average and 3-sigma concepts. It was finally confirmed that all three of the techniques detected the abnormal intervals correctly, while the weighted anomaly-detection technique is more precise than the basic techniques. These results show an excellent approach for the detection of log-data anomalies with the use of simple techniques in the Hadoop ecosystem.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2006.04a
• /
• pp.243-248
• /
• 2006

In this paper, we propose the unsupervised change detection algorithm that apply the similarity measure techniques to the hyperspectral image. The general similarity measures including euclidean distance and spectral angle were compared. The spectral similarity scale algorithm for reducing the problems of those techniques was studied and tested with Hyperion data. The thresholds for detecting the change area were estimated through EM(Expectation-Maximization) algorithm. The experimental result shows that the similarity measure techniques and EM algorithm can be applied effectively for the unsupervised change detection of the hyperspectral data.

• Structural Engineering and Mechanics
• /
• v.32 no.4
• /
• pp.489-500
• /
• 2009

The aim of this paper is to compare wavelet, kurtosis and pseudofractal based techniques for structural health monitoring in the presence of measurement noise. A detailed comparison and assessment of these techniques have been carried out in this paper through numerical experiments for the calibration of damage extent of a simply supported beam with an open crack serving as an illustrative example. The numerical experiments are deemed critical due to limited amount of experimental data available in the field of singularity based detection of damage. A continuous detectibility map has been proposed for comparing various techniques qualitatively. Efficiency surfaces have been constructed for wavelet, kurtosis and pseudofractal based calibration of damage extent as a function of damage location and measurement noise level. Levels of noise have been identified for each technique where a sudden drop of calibration efficiency is observed marking the onset of damage masking regime by measurement noise.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2013.10a
• /
• pp.757-762
• /
• 2013

A technique is presented that uses a new guided wave technique for sludge and blockages detection in long-range pipelines. Existing techniques have the limitations that the sludge position needs to be known a priori and the area to be inspected needs to be accessible. Two guided wave techniques have been developed which allow the sludge or blockages to be detected remotely without the need to access the specific location where the pipe is blocked, nor to open the pipe. The first technique measures the reflection of guided waves by sludge which can be used to accurately locate the blocked region; the second technique detects sludge by revealing the changes to the transmitted guided waves propagating in the blocked region or after it. The two techniques complement each other and their combination leads to a reliable sludge or blockage detection. Various types of realistic sludge have been considered in the study and the practical capabilities of the two techniques have been demonstrated.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.2781-2784
• /
• 2007

Early detection of a disease is important to tackle treatment issues in a better manner. Several diagnostic techniques are in use, these days; for such purpose and tremendous research is going on to develop newer and newer methods. However, more work is required to be done to develop cheap and reliable early detection techniques. Micro-fluidic chips are also playing key role to deliver new devices for better health care. The present study focuses on a review of recent developments in the interrogation of different techniques and present state-of-the-art of microfluidic sensor for better, quick, easy, rapid, early, inexpensive and portable POCT (Point of Care testing device) device for a particular study, in this case, bone disease called osteoporosis. Some simulations of the microchip are also made to enable feasibility of the development of a blood-chip-based system. The proposed device will assist in early detection of diseases in an effective and successful manner.