• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.6
• /
• pp.703-711
• /
• 2001

In this paper, we propose efficient multispectral image compression using variable block size vector quantization (VQ). In wavelet domain, we perform the variable block size VQ to remove intraband redundancy for a reference band image that has the lowest spatial variance and the best correlation with other band. And in wavelet domain, we perform the classified interband prediction to remove interband redundancy for the remaining bands. Then error wavelet coefficients between original image and predicted image are residual variable block size vector quantized to reduce prediction error. Experiments on remotely sensed satellite image show that coding efficiency of the proposed method is better than that of the conventional method.

• International Journal of Precision Engineering and Manufacturing
• /
• v.6 no.3
• /
• pp.37-44
• /
• 2005

Gait control capacity for most trans-femoral prostheses is significantly different from that of a normal person, and training is required for a long period of time in order for a patient to walk properly. People become easily tired when wearing a prosthesis or orthosis for a long period typically because the gait angle cannot be smoothly adjusted during wearing. Therefore, to improve the gait control problems of a trans-femoral prosthesis, the proper gait angle is estimated through surface EMG(electromyogram) signals on a normal leg, then the gait posture which the trans-femoral prosthesis should take is calculated in the neural network, which learns the gait kinetics on the basis of the normal leg's gait angle. Based on this predicted angle, a postural control method is proposed and tested adaptively following the patient's gait habit based on the predicted angle. In this study, the gait angle prediction showed accuracy of over $97\%$, and the posture control capacity of over $90\%$.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.12
• /
• pp.1755-1757
• /
• 2013

Dynamically exploiting unused-spectrum, cognitive radio has been proposed to solve spectrum utilization problem. In cognitive radio, it is important to minimize the interference to primary service as well as to provide efficient channel allocation. In this paper, we propose a multi-channel allocation scheme based on spectrum hole prediction. Proposed scheme considered both interference length and channel capacity to limit the interference to primary user as well as to enhance system performance. Simulation results show the proposed scheme improves the system throughput.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.85-88
• /
• 2001

Line spectrum pair(LSP) 계수는 양자화 오류에 강하고. 선형 릴간에 효율적이며, 필터의 안정성 판정이 용이하므로 LPC를 대신하여 음성 부호화에 널리 사용되고 있다. 일반적으로 LSP 계수간에는 일정한 상관관계가 나타나고, 이 특성을 이용하면 LSP 계수의 부호량을 줄일 수 있는 가능성이 있나. 본 논문에서는 LSP 계수를 압축하기 위해 principal component analysis(PCA)를 사용한 방법을 제안한다. 제안된 방법에서는 LSP 계수를 Karhunen-Loeve(KL) 변환해 에너지가 집중되는 고유치(eigenvalue)와 고유벡터(eigenvector)를 찾고 값을 양자화 한다. 성능 평가를 위해 2.4kbps MELP(mixed excitation linear prediction)와 8kbps QCELP(qualcumn code excited linear prediction) 음성 부호화기를 사용해 결과 값을 비교했고, 압축률이 증가하는 것을 확인했다.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.14 no.3
• /
• pp.241-245
• /
• 2001

In this paper, we present parameter optimization technique for GaAs/AlGaAs multiple quantum well avalanche photodiodes used for image capture mechanism in high-definition system. Even under flawless environment in semiconductor manufacturing process, random variation in process parameters can bring the fluctuation to device performance. The precise modeling for this variation is thus required for accurate prediction of device performance. The precise modeling for this variation is thus required for accurate prediction of device performance. This paper will first use experimental design and neural networks to model the nonlinear relationship between device process parameters and device performance parameters. The derived model was then put into genetic algorithms to acquire optimized device process parameters. From the optimized technique, we can predict device performance before high-volume manufacturign, and also increase production efficiency.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.22 no.9
• /
• pp.741-746
• /
• 2009

Compare to conventional Indium Tin Oxide (ITO) film deposition methods, cesium assisted sputtering method has been shown superior electrical, mechanical, and optical film properties. However, it is not easy to use cesium assisted sputtering method since ITO film properties are very sensitive to Cesium assisted equipment condition but their mechanism is not yet clearly defined physically or mathematically. Therefore, to optimize deposited ITO film characteristics, development of accurate and reliable process model is essential. For this, in this work, we developed ITO film deposition process model using neural networks and design of experiment (DOE). Developed model prediction results are compared with conventional statistical regression model and developed neural process model has been shown superior prediction results on modeling of ITO film thickness, sheet resistance, and transmittance characteristics.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1997.11a
• /
• pp.302-305
• /
• 1997

Treeing due to partial discharge(PD) is one of the main causes of breakdown of the insulating materials and reduction of tile insulation life. Therefore the necessity for establishing a method to diagnose the aging of insulation materials and to predict the breakdown of insulation has become important. From this viewpoint, our studies diagnose insulation degradation using the method of computer sensing system, which has the advantages of PD and acoustic emission(AE) sensing system. To use advantages of these two methods can be used effectively to search for treeing location and PD in some materials. In analysis method of degradation. using statically operator such as the center of gravity (G). the gradient of the discharge distribution(C), we have analyzed far tole prediction of life which we can be obtained the time, occurred of many pulse of small discharge amplitude.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2000.07a
• /
• pp.763-766
• /
• 2000

In this paper, stator form-winding sample coils based on silicone resin and polyimide were made for fault prediction and reliability estimation on the 200 Class insulation system of traction motors. The complex accelerative degradation was performed by periods during 10 cycles, which was composed of thermal stress, fast rising surge voltage, vibration, water immersion and overvoltage applying. After aging of 10 cycles, condition diagnosis test such as insulation resistance & polarization index, capacitance & dielectric loss and partial discharge properties were investigated in the temperature range of 20∼160$^{\circ}C$. Relationship among condition diagnosis test was analyzed to find an dominative degradation factor and an insulation state at end-life point.

• Proceedings of the KIEE Conference
• /
• 2005.07b
• /
• pp.942-944
• /
• 2005

In this study, eddy current signals from various anomalous defects in the heat exchanger tube are predicted af their signal slope characteristics no analyzed. The signal changes due to frequency increase are also observed. Based in the accumulated knowledge, the analysis of superimposed signal is attempted which includes the effects of support plate. Both differential and absolute bobbin probe signals are analyzed. For the prediction of signals, axisymmetric finite element modeling is used and this leads us to the utilization of slope angle analysis of the signal. Results show that differential signals are useful to locate the position of defect under the support plate and absolute signals no easy to predict and analyze even though they no superimposed signals. Combined use of these two types of signals will accomplish a reliable inspection.

• The Journal of Korea Robotics Society
• /
• v.19 no.2
• /
• pp.196-202
• /
• 2024

The ability to understand given environments and plan a sequence of actions leading to goal state is crucial for personal service robots. With recent advancements in deep learning, numerous studies have proposed methods for state representation in planning. However, previous works lack explicit information about relationships between objects when the state observation is converted to a single visual embedding containing all state information. In this paper, we introduce graph-based state representation that incorporates both object and relationship features. To leverage these advantages in addressing the task planning problem, we propose a Graph Neural Network (GNN)-based subgoal prediction model. This model can extract rich information about object and their interconnected relationships from given state graph. Moreover, a search-based algorithm is integrated with pre-trained subgoal prediction model and state transition module to explore diverse states and find proper sequence of subgoals. The proposed method is trained with synthetic task dataset collected in simulation environment, demonstrating a higher success rate with fewer additional searches compared to baseline methods.