• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.2
• /
• pp.391-399
• /
• 2012

This paper proposes a particle filter approach for SLAM(Simultaneous Localization and Mapping) of a mobile robot. The SLAM denotes estimation of both the robot location and map while the robot navigates in an unknown environment without map. The proposed method estimates robot location simultaneously with the locations of the ultrasonic beacons which constitute landmarks for navigation. The particle filter method represents the locations of the robot and landmarks in probabilistic manner by the distribution of particles. The method takes care of the uncertainty of the landmarks' location as well as that of the robot motion. Therefore, the locations of the landmarks are updated including uncertainty at every sampling time. Performance of the proposed method is verified through simulation and experiments. The method yields practically useful mapping information even if the range data from the landmarks include random noise. Also, it provides more accurate and robust estimation of the robot location than the usual least squares methods or dead-reckoning method.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.2
• /
• pp.85-89
• /
• 1996

In this paper, transition constrained Hidden Markov Model(HMM) in which the transition between states occur only within prescribed time slot is proposed and the performance is evaluated in the noisy environment. The transition constrained HMM can explicitly limit the state durations and accurately de scribe the temporal structure of speech signal simply and efficiently. The transition constrained HMM is not only superior to the conventional HMM but also require much less computation time. In order to evaluate the performance of the transition constrained HMM, speaker independent isolated word recognition experiments were conducted using semi-continuous HMM with the noisy speech for 20, 10, 0 dB SNR. Experiment results show that the proposed method is robust to the environmental noise. The 81.08% and 75.36% word recognition rates for conventional HMM was increased by 7.31% and 10.35%, respectively, by using transition constrained HMM when two kinds of noises are added with 10dB SNR.

• Journal of Korea Multimedia Society
• /
• v.3 no.4
• /
• pp.347-355
• /
• 2000

This paper presents an adaptive AEC(acoustic echo canceller) based on the wavelet transform using M-channel subband QMF filter banks. The proposed algorithm improves the performance of AEC with a realtime process by a low complexity of wavelet transform filter banks, a subband processing and a orthogonality of wavelet subband filter. Adaptive filter coefficients of each subband are updated using LMS algorithm with a low complexity and a easy realization for a realtime processing and a reduction of hardware cost. For a input signal, a white Gaussian noise and a real speech signal with a environment noises are used for a performance estimation of the proposed algorithm. As a result of computer simulation, the proposed AEC has a low asymptotic error, a low computation complexity and a robust performance.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.5 no.6
• /
• pp.584-591
• /
• 2000

Power Quality Compensator(PQC) has been installed to protect the sensitive loads against the voltage disturbances, such as voltage sag and interruption. In general, static switch is used for the purpose of link between utility and PQC. So transfer operation of the static switch play a important part in the PQC. Many studies on the structure and control of PQC have been progressed in active, but these researches have been rarely mentioned about any voltage-disturbances-detection method to start the PQC operation. In this paper, a new voltage-disturbances-detection algorithm for computer application loads using the CBEMA/ITIC curve is proposed for transfer operation of the static switch. The proposed detection algorithm is implemented to get fast detecting time through the comparison of instantaneous 3-phase voltage values transferred to DC values in the synchronous reference frame with the operating reference values. To get the robust characteristics against the noise, a first order digital filter is designed. The magnitude falling and phase delay caused by the filter are compensated through the error normalizing and numerical analysis using transfer function, respectively. Finally, the validity of the proposed algorithm is proved by ACSL simulation and experimental results.

• The KIPS Transactions:PartB
• /
• v.12B no.5 s.101
• /
• pp.587-594
• /
• 2005

This paper proposes a multi-level thresholding method for image segmentation using fuzzy clustering algorithm in transition region. Most of threshold-based image segmentation methods determine thresholds based on the histogram distribution of a given image. Therefore, the methods have difficulty in determining thresholds for real-image, which has a complex and undistinguished distribution, and demand much computational time and memory size. To solve these problems, we determine thresholds for real-image using fuzzy clustering algorithm after extracting transition region consisting of essential and important components in image. Transition region is extracted based on Inか entropy, which is robust to noise and is well-known as a tool that describes image information. And fuzzy clustering algorithm can determine optimal thresholds for real-image and be easily extended to multi-level thresholding. The experimental results demonstrate the effectiveness of the proposed method for performance.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.8
• /
• pp.2262-2270
• /
• 1999

In this paper, we describe gesture recognition algorithm using computer vision sensor and HMM. The automatic hand region extraction has been proposed for initializing the tracking of the tele-operation gestures. For this, distance informations(disparity map) as results of stereo matching of initial left and right images are employed to isolate the hand region from a scene. PDOE(positive difference of edges) feature images adapted here have been found to be robust against noise and background brightness. The KNU/KAERI(K/K) gesture instruction set is defined for tele-operation in atomic electric power stations. The composite recognition model constructed by concatenating three gesture instruction models including pre-orders, basic orders, and post-orders has been proposed and identified by discrete HMM. Our experimental results showed that consecutive orders composed of more than two ones are correctly recognized at the rate of above 97%.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4C
• /
• pp.414-423
• /
• 2009

This paper considers a linear precoding scheme that achieves near optimal sum rate. While the minimum mean square error (MMSE) precoding provides the better MSE performance at all signal-to-noise ratio (SNR) than the zero forcing (ZF) precoding, its sum rate shows superior performance to ZF precoding at low SNR but inferior performance to ZF precoding at high SNR, From this observation, we first propose a near optimal linear precoding scheme in terms of sum rate. The resulting precoding scheme regularizes ZF precoding to maximize the sum rate, resulting in better sum rate performance than both ZF precoding and MMSE precoding at all SNR ranges. To find regularization parameters, we propose a simple algorithm such that locally maximal sum rate is achieved. As a low complexity alternative, we also propose a simple power re-allocation scheme in the conventional regularized channel inversion scheme. Finally, the proposed scheme is tested under the presence of channel estimation error. By simulation, we show that the proposed scheme can maintain the performance gain in the presence of channel estimation error and is robust to the channel estimation error.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.12
• /
• pp.573-580
• /
• 2017

Deep Neural Network (DNN) is a large layered neural network which is consisted of a number of layers of non-linear units. Deep Learning which represented as DNN has been applied very successfully in various applications. However, many issues in DNN have been identified through past researches. Among these issues, generalization is the most well-known problem. A Recent study, Dropout, successfully addressed this problem. Also, Dropout plays a role as noise, and so it helps to learn robust feature during learning in DNN such as Denoising AutoEncoder. However, because of a large computations required in Dropout, training takes a lot of time. Since Dropout keeps changing an inter-layer representation during the training session, the learning rates should be small, which makes training time longer. In this paper, using mutation operation, we reduce computation and improve generalization performance compared with Dropout. Also, we experimented proposed method to compare with Dropout method and showed that our method is superior to the Dropout one.

• Journal of the Institute of Convergence Signal Processing
• /
• v.4 no.1
• /
• pp.1-8
• /
• 2003

Various methods have been studying to maintain and apply the multimedia inform abruptly increasing over all social fields, in recent years. For retrieval of still images, we is implemented content-based image retrieval system in this paper that make possible to retrieve similar objects from image database after segmenting query object from background if user request query. Query image is processed median filtering to remove noise first and then object edge is detected it by canny edge detection. And query object is segmented from background by using convex hull. Similarity value can be obtained by means of histogram intersection with database image after securing color histogram from segmented image. Also segmented image is processed gray convert and wavelet transform to extract spacial gray distribution and texture feature. After that, Similarity value can be obtained by means of banded autocorrelogram and energy. Final similar image can be retrieved by adding upper similarity values that it make possible to not only robust in background but also better correct object retrieval by using object segmentation method.

• Journal of Broadcast Engineering
• /
• v.18 no.6
• /
• pp.850-858
• /
• 2013

In this paper, we propose a method of frame synchronization for stereoscopic 3D video to solve the viewing problem caused by synchronization errors between a left video and a right video using the temporal frame difference image depending on the movement of objects. Firstly, we compute two temporal frame difference images from the left video and the right video which are corrected the vertical parallax between two videos using rectification, and calculate two horizontal projection profiles of two temporal frame difference images. Then, we find a pair of synchronized frames of the two videos by measuring the mean of absolute difference (MAD) of two horizontal projection profiles. Experimental results show that the proposed method can be used for stereoscopic 3D video, and is robust against Gaussian noise and video compression by H.264/AVC.