• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.7
• /
• pp.70-82
• /
• 1993

In this paper, employing the 3-dimensional discrete cosine transform (DCT) for the utilization of the temporal correlation, an adaptive motion sequence coding technique is proposed. The energy distribution in a 3-D DCT block, due to the nonstationary nature of the image data, varies along the veritical, horizontal and temporal directions. Thus, aiming an adaptive system to local variations, adaptive procedures, such as the 3-D classification, the classified linear scanning technique and the VLC table selection scheme, have been implemented in our approach. Also, a hybrid structure which adaptively combines inter-frame coding is presented, and it is found that the adaptive hybrid frame coding technique shows a significant performance gain for a moving sequence which contains a relatively small moving area. Through an intensive computer simulation, it is demonstrated that, the performance of the proposed 3-D transform coding technique shows a close relation with the temporal variation of the sequence to be code. And the proposed technique has the advantages of skipping the computationally complex motion compensation procedure and improving the performance over the 2-D motion compensated transform coding technique for rates in the range of 0.5 ~ 1.0 bpp.

• International Journal of Automotive Technology
• /
• v.8 no.5
• /
• pp.637-650
• /
• 2007

A new on-line fault detection and isolation(FDI) scheme has been proposed for engines using an adaptive neural network classifier; this paper investigates the robustness of this scheme by evaluating in a wide range of operational modes. The neural classifier is made adaptive to cope with the significant parameter uncertainty, disturbances, and environmental changes. The developed scheme is capable of diagnosing faults in the on-line mode and can be directly implemented in an on-board diagnosis system(hardware). The robustness of the FDI for the closed-loop system with crankshaft speed feedback is investigated by testing it for a wide range of operational modes, including robustness against fixed and sinusoidal throttle angle inputs, change in load, change in an engine parameter, and all changes occurring simultaneously. The evaluations are performed using a mean value engine model(MVEM), which is a widely used benchmark model for engine control system and FDI system design. The simulation results confirm the robustness of the proposed method for various uncertainties and disturbances.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.4
• /
• pp.399-407
• /
• 2002

In this paper, a novel algorithm is proposed to reduce the blocking artifacts of block-based coded images by using block classification and MLP. In the proposed algorithm, we classify the block into four classes based on a characteristic of DCT coefficients. And then, according to the class information of neighborhood block, adaptive neural network filter is performed in horizontal and vertical block boundary. That is, for smooth region, horizontal edge region, vertical edge region, and complex region, we use a different two-layer neural network filter to remove blocking artifacts. Experimental results show that the proposed algorithm gives better results than the conventional algorithms both subjectively and objectively.

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.11
• /
• pp.521-528
• /
• 2015

Road-view object classification methods are mostly influenced by weather and illumination conditions, thus the most of the research activities are based on dataset in clean weathers. In this paper, we present a road-view object classification method based on color segmentation that works for all kinds of weathers. The proposed method first classifies the weather and illumination conditions and then applies the weather-specified color models to find the road traffic signs. Using 5 different features of the road-view images, we classify the weather and light conditions as sunny, cloudy, rainy, night, and backlight. Based on the classified weather and illuminations, our model selects the weather-specific color ranges to generate Gaussian Mixture Model for each colors, Green, Yellow, and Blue. The proposed method successfully detects the traffic signs regardless of the weather and illumination conditions.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1590-1599
• /
• 2016

The installation of distributed generation (DG) in the electrical networks has numerous advantages. However, connecting and disconnecting of DGs (CADD) leads to some problems in coordination of protection devices due to the changes in the short circuit levels in the different points of network. In this paper, an adaptive method is proposed based on available setting groups (SG) of relays. Since the number of available SG is less than possible CADD states, a classifying index (CI) is defined to categorize the several states in restricted setting groups. Genetic algorithm (GA) with a suitable objective function (OF) is used as an optimization method for the classification. After grouping, a modified coordination method is applied to achieve optimal coordination for each group. The efficiency of the proposed technique is demonstrated by simulation results.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.3E
• /
• pp.124-132
• /
• 2003

We propose a new method for speech/non-speech classifiers based on concepts of the adaptive boosting (AdaBoost) algorithm in order to detect speech for robust speech recognition. The method uses a combination of simple base classifiers through the AdaBoost algorithm and a set of optimized speech features combined with spectral subtraction. The key benefits of this method are the simple implementation, low computational complexity and the avoidance of the over-fitting problem. We checked the validity of the method by comparing its performance with the speech/non-speech classifier used in a standard voice activity detector. For speech recognition purpose, additional performance improvements were achieved by the adoption of new features including speech band energies and MFCC-based spectral distortion. For the same false alarm rate, the method reduced 20-50% of miss errors.

• Proceedings of the IEEK Conference
• /
• 1998.06a
• /
• pp.677-680
• /
• 1998

In this paper, we proposed a lwo bit rate subband coding with adaptive vector quantization using the correlation between motion vector and block energy in subband. In this method, the difference between the input signal and the motion compensated interframe prediction signal is decomposed into several narrow bands using quadrature mirror filter (QMF) structure. The subband signals are then quantized by adaptive vector quantizers. In the codebook generating process, each classified region closer to the block value in the same region after the classification of region by the magnitude of motion vector and the variance values of subband block. Because codebook is genrated considering energy distribution of each region classified by motion vector and variance of subband block, this technique gives a very good visual quality at low bit rate coding.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.10
• /
• pp.1546-1552
• /
• 1993

We proposed an adaptive block truncation coding(BTC) using gradient and a new initial value. We used gradient of sobel operator as a new category classification coefficient to reduce Jagged appearance at edge part. We defined a new initial value to reduce large quantization error in the 4-level quantizer block including edge part. By computer simulations, we showed that the proposed method less computation load, reduced jagged appearance at edge part, also improved PSNR more than the conventional adaptive BTC.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.971-972
• /
• 2008

In this paper, we present a novel method for classifying pornography images using adaptive skin detection. From an input image, we detect initial skin regions and construct an adaptive skin probability density model using color information for the detected skin regions. From the skin probability density model, we extract feature vectors and train the images using Support Vector Machine to classify pornography images.

• Proceedings of the Korea Electromagnetic Engineering Society Conference
• /
• 2002.11a
• /
• pp.44-48
• /
• 2002

This paper describes an adaptive signal processing of digital receiver with DDC(Digital Down Convertor), DDC is implemented by using NCO(Numerically Controlled Oscillator), digital low pass filter. for the passband sampling, we present the results of digital receiver simulation with DDC. We confirm that the low IP signal is converted to zero IF by DDC. DOA(Direction Of Arrival) estimation technique using MUSIC(Multiple SIgnal Classification) algorithm with high resolution is presented. We Cow that an accurate resolution of DOA depends on the input sampling number.