• Journal of Korea Multimedia Society
• /
• v.7 no.2
• /
• pp.156-163
• /
• 2004

In this paper, we propose an adaptive face region detection and real-time face identification algorithm using face feature evaluation function. The proposed algorithm can detect exact face region adaptively by using skin color information for races as well as intensity and elliptical masking method. And also, it improves face recognition efficiency using geometrical face feature and geometric evaluation function between features. The proposed algorithm can be used for the development of biometric and security system areas. In the experiment, the superiority of the proposed method has been tested using real image, the proposed algorithm shows more improved recognition efficiency as well as face region detection efficiency than conventional method.

• Smart Structures and Systems
• /
• v.12 no.6
• /
• pp.619-640
• /
• 2013

It is highly desirable to explore efficient algorithms for detecting structural damage of large size structural systems with limited input and output measurements. In this paper, a new structural damage detection algorithm based on substructure approach is proposed for large size structural systems with limited input and output measurements. Inter-connection effect between adjacent substructures is treated as 'additional unknown inputs' to substructures. Extended state vector of each substructure and its unknown excitations are estimated by sequential extended Kalman estimator and least-squares estimation, respectively. It is shown that the 'additional unknown inputs' can be estimated by the algorithm without the measurements on the substructure interface DOFs, which is superior to previous substructural identification approaches. Also, structural parameters and unknown excitation are estimated in a sequential manner, which simplifies the identification problem compared with other existing work. Structural damage can be detected from the degradation of the identified substructural element stiffness values. The performances of the proposed algorithm are demonstrated by several numerical examples and a lab experiment. Measurement noise effect is considered. Both the simulation results and experimental data validate that the proposed algorithm is viable for structural damage detection of large size structural systems with limited input and output measurements.

• Aerospace Engineering and Technology
• /
• v.6 no.2
• /
• pp.21-28
• /
• 2007

The goal of this paper is to represent a technique of fault detection for the control surface as a base research of the fault tolerant control system for safety improvement of UAV. The real-time system identification based on the recursive Fourier Transform was implemented for the fault detection of the control surface and verified through the HILS and flight test. The failures of the control surface are detected by comparing the control derivatives in fault condition with the normal condition. As a result from the flight test, we have confirmed that the control derivatives of fault condition less than normal condition.

• Nuclear Engineering and Technology
• /
• v.17 no.2
• /
• pp.83-97
• /
• 1985

The Generalized Likelihood Ratio (GLR) method performs statistical tests on the innovations sequence of a Kalman-Buchy filter state estimator for system failure detection and its identification. However, the major drawback of the conventional GLR is to hypothesize particular failure type in each case. In this paper, a method to solve this drawback is proposed. The improved GLR method is applied to a PWR pressurizer and gives successful results in detection and identification of any failure. Furthermore, some benefit on the processing time lier each cycle of failure detection and its identification can be accompanied.

• Journal of Information Processing Systems
• /
• v.18 no.5
• /
• pp.677-687
• /
• 2022

We propose a detection algorithm based on tree-structured convolutional neural networks (TsCNNs) that finds pornography, propaganda, or other inappropriate content on a social media network. The algorithm sequentially applies the typical convolutional neural network (CNN) algorithm in a tree-like structure to minimize classification errors in similar classes, and thus improves accuracy. We implemented the detection system and conducted experiments on a data set comprised of 6 ordinary classes and 11 inappropriate classes collected from the Korean military social network. Each model of the proposed algorithm was trained, and the performance was then evaluated according to the images and videos identified. Experimental results with 20,005 new images showed that the overall accuracy in image identification achieved a high-performance level of 99.51%, and the effectiveness of the algorithm reduced identification errors by the typical CNN algorithm by 64.87 %. By reducing false alarms in video identification from the domain, the TsCNNs achieved optimal performance of 98.11% when using 10 minutes frame-sampling intervals. This indicates that classification through proper sampling contributes to the reduction of computational burden and false alarms.

• Korean Journal of Remote Sensing
• /
• v.40 no.2
• /
• pp.167-177
• /
• 2024

With advancements in satellite technology, interest in target detection and identification is increasing quantitatively and qualitatively. Synthetic Aperture Radar(SAR) images, which can be acquired regardless of weather conditions, have been applied to various areas combined with machine learning based detection algorithms. However, conventional studies primarily focused on the detection of stationary targets. In this study, we proposed a method to identify moving targets using an algorithm that integrates sub-aperture SAR images and cosine similarity calculations. Utilizing a transformer-based deep learning target detection model, we extracted the bounding box of each target, designated the area as a region of interest (ROI), estimated the similarity between sub-aperture SAR images, and determined movement based on a predefined similarity threshold. Through the proposed algorithm, the quantitative evaluation of target identification capability enhanced its accuracy compared to when training with the targets with two different classes. It signified the effectiveness of our approach in maintaining accuracy while reliably discerning whether a target is in motion.

• Journal of Microbiology and Biotechnology
• /
• v.34 no.3
• /
• pp.681-688
• /
• 2024

The accurate and rapid detection of methicillin-resistance of Staphylococcus aureus (SA) holds significant clinical importance. However, the methicillin-resistance detection strategies commonly require complicated cell lysis and gene extraction. Herein, we devised a novel colorimetric approach for the sensitive and accurate identification of methicillin-resistance of SA by combining allosteric probe-based target recognition with self-primer elongation-based target recycling. The PBP2a aptamer in the allosteric probe successfully identified the target MRSA, leading to the initiation of self-primer elongation based-cascade signal amplification. The peroxidase-like hemin/G-quadruplex undergo an isothermal autonomous process that effectively catalyzes the oxidation of ABTS2- and produces a distinct blue color, enabling the visual identification of MRSA at low concentrations. The method offers a shorter duration for bacteria cultivation compared to traditional susceptibility testing methods, as well as simplified manual procedures for gene analysis. The overall amplification time for this test is 60 min, and it has a detection limit of 3 CFU/ml. In addition, the approach has exceptional selectivity and reproducibility, demonstrating commendable performance when tested with real samples. Due to its advantages, this colorimetric assay exhibits considerable potential for integration into a sensor kit, thereby offering a viable and convenient alternative for the prompt and on-site detection of MRSA in patients with skin and soft tissue infections.

• Structural Monitoring and Maintenance
• /
• v.1 no.4
• /
• pp.411-425
• /
• 2014

A relatively low frequency Lamb wave-based damage identification method called damage imaging method for rectangular composite plate is presented. A damage index (DI) is generated from the delay matrix of the Lamb wave response signals, and it is used to indicate the location and approximate area of the damage. The viability of this method is demonstrated by analyzing the numerical and experimental Lamb wave response signals from rectangular composite plates. The technique only requires the response signals from the plate after damage, and it is capable of performing near real time damage identification. This study sheds some light on the application of Lamb wave-based damage detection algorithm for plate-type structures by using the relatively low frequency (e.g., in the neighborhood of 100 kHz, more suitable for the best capability of the existing fiber optic sensor interrogator system with the sampling frequency of 500 kHz) Lamb wave response and a reference-free damage detection technique.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.12
• /
• pp.1667-1670
• /
• 2014

This paper describes a method for detecting the pupil region and identification of the eye status for driver drowsiness detection system. This program detects a driver's face and eyes using viola-jones face detection algorithm and extracts the pupil area by utilizing mean values of each row and column on the eye area. The proposed method uses binary images and the number of black pixels to identify the eye status. Experimental results showed that the accuracy of classification eye status(open/close) was above 90%.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.66 no.1
• /
• pp.27-32
• /
• 2017

This paper presents a Daubechies wavelet-based fault detection method for fault identification in transmission lines. After the Daubechies wavelet coefficients are calculated, the proposed algorithm has been implemented difference equation using C language. We have modeled a 154kV transmission line using the ATPDraw software and have acquired test data. In order to evaluate effects of DC offset, simulations carried out while varying an inception angle of the voltage $0^{\circ}$, $45^{\circ}$, $90^{\circ}$. For performance evaluation, fault distance was varied. As we can see from the off-line simulation, the proposed algorithm shows rapid and accurate fault detection. Also we can see the proposed algorithm is not affected by the fault inception angle change.