• KIISE Transactions on Computing Practices
• /
• v.23 no.7
• /
• pp.408-416
• /
• 2017

This paper proposes a real-time side-rear vehicle detection algorithm that detects vehicles quickly and accurately in blind spot areas when driving. The proposed algorithm uses a cascade classifier created by AdaBoost Learning using the MCT (modified census transformation) feature vector. Using this classifier, the smaller the detection window, the faster the processing speed of the MCT classifier, and the larger the detection window, the greater the accuracy of the MCT classifier. By considering these characteristics, the proposed algorithm uses two classifiers with different detection window sizes. The first classifier quickly generates candidates with a small detection window. The second classifier accurately verifies the generated candidates with a large detection window. Furthermore, the vehicle classifier and the wheel classifier are simultaneously used to effectively detect a vehicle entering the blind spot area, along with an adjacent vehicle in the blind spot area.

• Journal of Korean Society for Atmospheric Environment
• /
• v.25 no.1
• /
• pp.46-56
• /
• 2009

The main purpose of this study was to investigate air quality trends of ambient aerosol with obtaining size-fractionated information. The suspended particulate matters were continuously collected on membrane filters and glass fiber filters by an 8-stage cascade impactor for 2 years (Sep. 2005 $\sim$ Sep. 2007) in Kyung Hee University-Global Campus. 8 ionic species ($Na^+$, ${NH_4}^+$, $K^+$, $Mg^{2+}$, $Ca^{2+}$, $Cl^-$, ${NO_3}^-$, and ${SO_4}^{2-}$) were analyzed by an IC after performing proper pretreatments of each sample filter. The average concentration levels of each ion were $9.24{\mu}g/m^3$ of ${SO_4}^{2-}$, $7.35{\mu}g/m^3$ of ${NO_3}^-$, $2.81{\mu}g/m^3$ of ${NH_4}^+$, $2.11{\mu}g/m^3$ of $Ca^{2+}$, $1.65{\mu}g/m^3$ of $Cl^-$, $1.87{\mu}g/m^3$ of $Na^+$, $0.80{\mu}g/m^3$ of $Mg^{2+}$, and $0.54{\mu}g/m^3$ of $K^+$, respectively. The distribution pattern of $Na^+$, $Mg^{2+}$, $Ca^{2+}$, $Cl^-$, and ${NO_3}^-$ was bi-modal and two peaks appeared in the range of $0.4{\sim}0.7{\mu}m$ and $3.3{\sim}4.7{\mu}m$, respectively. On the other hand, ${SO_4}^{2-}$, ${NH_4}^+$, and $K^+$ showed patterns of uni-modal distribution, mostly abounded in the fine mode group.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.10
• /
• pp.166-174
• /
• 1998

Neural networks, having highly nonlinear dynamics by virtue of the distributed nonlinearities and the learing ability, have the potential for the adaptive prediction of nonstationary signals. This paper describes the nonlinear prediction of these signals in two ways; using a nonlinear module and the cascade combination of nonlinear and linear modules. Fully-connected recurrent neural networks (RNNs) and a conventional tapped-delay-line (TDL) filter are used as the nonlinear and linear modules respectively. The dynamic behavior of the proposed predictors is demonstrated for chaotic time series adn speech signals. For the relative comparison of prediction performance, the proposed predictors are compared with a conventional ARMA linear prediction model. Experimental results show that the neural networks based adaptive predictor ourperforms the traditional linear scheme significantly. We also find that the cascade combination predictor is well suitable for the prediction of the time series which contain large variations of signal amplitude.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.19 no.5
• /
• pp.813-827
• /
• 2017

Surveillance cameras installed in tunnels capture the various video frames effected by dynamic and variable factors. In addition, localizing and managing the cameras in tunnel is not affordable, and quality of capturing frame is effected by time. In this paper, we introduce a new method to detect and track the vehicles in tunnel by using surveillance cameras installed in a tunnel. It is difficult to detect the video frames directly from surveillance cameras due to the motion blur effect and blurring effect on lens by dirt. In order to overcome this difficulties, two new methods such as Differential Frame/Non-Maxima Suppression (DFNMS) and Haar Cascade Detector to track cars are proposed and investigated for their feasibilities. In the study, it was shown that high precision and recall values could be achieved by the two methods, which then be capable of providing practical data and key information to an automatic accident detection system in tunnels.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.13 no.5
• /
• pp.923-930
• /
• 2018

In the wireless communication system based on an array antenna, the angle of arrival (AOA) information of signal is very important element and various AOA estimation algorithms have been studied. Although most AOA estimation algorithms employ the uniform linear array (ULA), some algorithms apply the planar array (PA) antenna. In this paper, we present an algorithm for efficiently estimating AOAs of adjacent multiple signals, based on the uniform rectangular array antenna. This approach has two steps; after approximately estimating AOA groups consisting of the closely located signal sources using CAPON, accurately estimating the individual AOA of each signal in the estimated AOA group using Beamsapce MUSIC. The estimation performance of the presented cascade AOA algorithm is illustrated through the computer simulation example.

• Journal of Convergence for Information Technology
• /
• v.12 no.3
• /
• pp.294-301
• /
• 2022

In this study, a multi-layer cascade (MLC) electrical array configuration method for thermoelectric generator consisting of plural number of thermoelectric modules (TEMs) was proposed to reduce mismatch loss caused by temperature maldistribution on the surfaces of the TEMs. To validate the effect of MLC on the mismatch loss reduction, a numerical model capable of reflecting multi-physics phenomena occuring in the TEMs was developed. MLC can be employed by placing a group of TEMs experiencing relatively low temperature differences in an electric layer with more electrical branches while locating a group of TEMs experiencing relatively high temperature differences in an electric layer with less electrical branches. The TEMs were classified using the temperature distribution obtained by the numerical model. A MLC with an optimal electrical branch ratio showed a 96.5% of electric power generation compared to an ideal case.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.955-959
• /
• 2003

A technique for realizing linear phase IIR filters has been proposed by Powell-Chau which gives a real-time implementation of H(z-1).H(z), where H(z) is a causal nonlinear phase IIR filter. Powell-Chau system is linear but not timeinvariant system. Therefore, that system has group delay response that exhibits a minor sinusoidal variation superimposed on a constant value. In the signal processing, this oscillation seriously degrade the signal quality. Unfortunately, that system has a large sample delay of 4L and also more computational complexity. Proposed system is present a reduced computational complexity technique by moved the numerator polynomial of H(1/z) out to cascade with causal filter H(z) and remain only all-pole of H(1/z), then applied truncated infinite impulse response to finite with truncated IIR filtel $H_L$(z) and L sample delay to subtract the output sequence from the top and bottom filter. Proposed system is linear time invariance and group delay response and total harmonic distortion are also improved.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.10
• /
• pp.4117-4135
• /
• 2020

Due to the angle of view, illumination and scene diversity, real-time dynamic face detection and recognition is no small difficulty in those unrestricted environments. In this study, we used the intrinsic correlation between detection and calibration, using a multi-task cascaded convolutional neural network(MTCNN) to improve the efficiency of face recognition, and the output of each core network is mapped in parallel to a compact Euclidean space, where distance represents the similarity of facial features, so that the target face can be identified as quickly as possible, without waiting for all network iteration calculations to complete the recognition results. And after the angle of the target face and the illumination change, the correlation between the recognition results can be well obtained. In the actual application scenario, we use a multi-camera real-time monitoring system to perform face matching and recognition using successive frames acquired from different angles. The effectiveness of the method was verified by several real-time monitoring experiments, and good results were obtained.

• Journal of Power Electronics
• /
• v.17 no.3
• /
• pp.704-715
• /
• 2017

For three-phase synchronization applications, the synchronous reference frame phase-locked loop (SRF-PLL) is probably the most widely used technique due to its ease of implementation and satisfactory phase tracking performance under ideal grid conditions. However, under unbalanced and distorted grid conditions, its performance tends to worsen. To deal with this problem, a variety of filtering stages have been proposed and used in SRF-PLLs for the rejection of disturbance components at the cost of degrading the dynamic performance. In this paper, to improve dynamic performance without compromising the filtering capability, an effective hybrid filtering stage is proposed and incorporated into the inner loop of a quasi-type-1 PLL (QT1-PLL). The proposed filtering stage is a combination of a moving average filter (MAF) and a modified delay signal cancellation (DSC) operator in cascade. The time delay caused by the proposed filtering stage is smaller than that in the conventional MAF-based and DSC-based PLLs. A small-signal model of the proposed PLL is derived. The stability is analyzed and parameters design guidelines are given. The effectiveness of the proposed PLL is confirmed through experimental results.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.6
• /
• pp.1869-1880
• /
• 2014

Wireless communication is one of most active areas in modern communication researches, QoS (Quality of Service) assurance is very important for wireless communication systems design, especially for applications such as streaming video etc., which requires strict QoS assurance. The modern wireless networks multi-layer protocol stack structure results in QoS metrics layered and acting in cascade and QoS metrics vertical mapping between protocol layers. Based on virtual buffer between protocol layers and queuing technology, a unified layered QoS mapping framework is proposed in this paper, in which we first propose virtual queue concept, give a novelty united neighboring protocol layers QoS metric mapping framework, and analysis method based on dicerete-time Markov chain, and numerical results show that our proposed framework represents a significant improvement over previous model.