The paper handles a skin tone enhancement within digitalized images, which is based on human favorite skin color. To archive a human-favorite skin tone area, lots of observers are participated in visual experiments. The statistical result of the visual experiment was reflected in a proposed skin tone enhancement algorithm and gives more reasonable human-friendly skin color reproduction on a display device. This research was accomplished to implement in our DTV related application.

In this Paper, we propose an efficient de-interlacing algorithm using temporal and spatial domain information. In the proposed scheme, motion estimation is performed same parity fields, i.e., if current field is even field, reference fields are previous even field and forward even field. And then motion vector refinement is performed to improve the accuracy of motion vectors. In the interpolating step, we use median filter to reduce the interpolation error caused by incorrect motion vector. Simulations conducted for various video sequences have shown the efficiency of the proposed interpolator with significant improvement over previous methods in terms of both PSNR and perceived image quality.

This paper presents a new baseband modulation system for the reception improvement of VSB system in dynamic multipath channel. By introducing the 1/2 Trellis Coded Modulation in baseband modulation, the TOV of proposed modulation is improved very much compared with 8-VSB modulation and better than the pseudo 2-VSB modulation of Philips which is the most robust modulation for VSB transmission system as far. Through the computer simulation, the performance of the proposed modulation is examined and the results are shown.

The methods that track and detect motion field based on image difference of successive images from camera can separate motion field and background effectively, but because of noise and background images getting proper difference images is hard to achieve. In this paper we propose a method that can improve difference image quality significantly. Three step process is used. At the first step, existence of motion field is determined, the second step is finding proper threshold value using 'Contrast Streching' technique which enables us to find proper motion field even in complex images. At last step, remaining noise is removed and motion field is determined.

This paper presents a new method for the motion detection of vehicles using hierarchical motion estimation and parallel processing. It captures the road image by using a CMOS sensor. It divides the captured image into small blocks and detects the motion of each block by using a block-matching method which is based on a hierarchical motion estimation and parallel processing for the real-time processing. The parallelism is achieved by using the pipeline and the data flow technique. The proposed method has been implemented with an embedded system. Experimental results show that the proposed method detects the motion of vehicles in real-time.

This paper presents variable bit rate(VBR) rate control scheme based on MPEG-4 VM8 rate control scheme. An initial Q searching method provides more accurate bit allocation for the first frame. A frame skipping and RD Model update scheme when coded frame quality is too low or high prevents image quality fluctuation.

In typical block-based video coding, the objective of RC(Rate Control) is to select the quantization parameters so that the encoder produces bits at the rate of the channel and the overall distortion is minimized. To reduce the huge amount of computations required for offline RC, there have been significant efforts to speed up the process of video encoders. Those efforts have been mainly focused on the modes for bit rate and distortion in types of coders, in terms of the quantization parameters. Because previous works related to model based online RC are based on statistics of previous frame, it occurs the problem such that allocates bits unequally without regard to current frame statistics. In this thesis, an equal bit allocation scheme using current frame statistics is proposed.

Multiple Description Coding(MDC) is a technique used to obtain two or more (often correlated) descriptions of a source, which are transmitted over different channels to receiver. Two descriptions of the source support two levels of reconstruction quality. When all the descriptions are received and used in the reconstruction, the source should be reconstructed with acceptable quality. In this work, we consider Multiple Description Scalar Quantizer(MDSQ) to wavelet transform domain. Conventional MDSQ schemes in wavelet domain considered description with equal weights at each sub-bands after quantization. But each sub-bands is unequal contribution to whole image quality. Therefore, we experiment the multiple description with unequal weight in each sub-bands.

MPEG-4 부호화기는 16×16 픽셀로 이루어진 매크로 블록을 기본 단위로 하는 비디오 객체에 대해 움직임 추정과 움직임 보상을 수행한다. 이때 이전 블록과 현재 블록간의 차이를 8×8 블록으로 BCT하여 압축한다. DCT는 영상 신호의 에너지를 낮은 대역으로 집중시켜 낮은 에너지를 갖는 고주파 성분을 제거하여 정보를 압축하므로 복원된 영상은 블록의 경계선이 표시되는 블록킹 현상이 발생한다. 본 논문에서는 4×4 블록과 8×8 블록을 이용하여 DCT하는 가변 블록 변환 방식을 사용한다. 이는 작은 크기의 블록을 사용하므로 블록내 주파수의 분포를 비교적 좁은 영역으로 변환할 수 있어 블록킹 현상을 줄일 수 있다. 또한 INTRA 프레임의 DCT 과정에서 4개의 DCT 변환 블록중 하나를 이웃하는 블록의 DC 계수와 화소값들을 이용하여 그 값을 예측하여 DCT계수를 결정하고 INTER 프레임에서는 간략화된 정수 DCT의 사용과 작은 블록단위의 작업으로 인한 부호화의 부담을 줄인 다.

In the communications systems with diversity, we are commonly faced on needing of new source coding technique, error resilient coding. The error resilient coding addresses the coding algorithm that has the robustness to unreliability of communications channel. In recent years, many error resilient coding techniques were proposed such as data partitioning, resynchronization, error detection, concealment, reference picture selection and multiple description coding (MDC). In this paper, we proposed an MDC using whitening transform. The conventional MDC using correlating transform is need additional information to decode the image. But, if an image is transformed using the whitening transform, the additional information is not necessary to transform because the coefficients of whitening transform have uni-variance statistics.

In this paper, we propose an iris recognition system using Homogeneous Texture descriptor of MPEG-7 standard. The texture of iris is generally used in iris recognition system. We segment the pupil with Hough transform and the boundary of iris with it's gray level difference between the white of the eye. To extract Homogeneous Texture descriptor, this iris image is transformed into polar coordinates. The extracted descriptor is then compared with the reference in DB. If their distance is larger than threshold, they are recognized as different iris. Test results will show that Homogeneous Texture descriptor can be a good measure for iris recognition system.

This paper presents a fast algorithm for face detection from color images on internet. We use Mahalanobis distance between standard skin color and actual pixel color on IQ color space to segment skin color regions. The skin color regions are the candidate face region. Further, the locations of eyes and mouth regions are found by computing average pixel values on horizontal and vertical pixel lines. The geometry of mouth and eye locations is compared to the standard face geometry to eliminate false face regions. Our Method is simple and fast so that it can be applied to face search engine for internet.

In this paper, we propose the method of iris lacuna extraction using watershed transform. Lacuna is salient feature of iris. It has three dimensional structure formed by leak of pigmentation and loss of fiber tissues. Lacuna can be used for iris recognition system, and generally used in health diagnosis and character analysis with its shape and position. The main idea of the proposed method is applying the watershed transform to radial gray scale profile of iris image. The result shows that the lacuna can be extracted automatically from eye image.

Facial features are often used for human computer interface(HCI). This paper proposes a method to detect facial features using color and facial geometry information. Face region is first extracted by using color information, and then the pupils are detected by applying a separability filter and facial geometry constraints. Mouth is also extracted from Cr(coded red) component. Experimental results shows that the proposed detection method is robust to a wide range of facial variation in position, scale, color and gaze.

In this paper, we propose the iris feature extraction by using statistical methods. There are many approaches for iris feature extraction, but most of them require a set of parameters that one should choose for the transformation to obtain a useful representation of the iris. It would be most useful to estimate the method of the iris feature extraction from iris itself. Therefore, we apply the unsupervised statistical methods for the iris feature extraction.

In an automatic fingerprint recognition system, a thinning process after binarization is commonly used. However it gives rise to spurs and holes often causing many spurious minutiae. Thus, more elaborate postprocessing is urgently needed to remove such spurious minutiae. To overcome this problem, we present a method of extracting minutiae based on horizontal and vertical run-length encoding from a binary fingerprint image without thinning process. Experimental results show that the proposed method for extracting minutiae is fairly reliable and fast, when il is compared to other method adopting a thinning process.

Human face detection has many applications such as face recognition, face or facial feature tracking, pose estimation, and expression recognition. We present a new method for automatically segmentation and face detection in color images. Skin color alone is usually not sufficient to detect face, so we combine the color segmentation and shape analysis. The algorithm consists of two stages. First, skin color regions are segmented based on the chrominance component of the input image. Then regions with elliptical shape are selected as face hypotheses. They are certificated to searching for the facial features in their interior, Experimental results demonstrate successful detection over a wide variety of facial variations in scale, rotation, pose, lighting conditions.

This paper has been studied an algorithm for segmenting gaseous object images. A new methodology for segmenting gaseous object images is introduced. Proposed method consists offuzzy-based boundary detector applicable to gaseous as well as rigid objects and concave region filling to recover object regions.

In this paper, We have studied for guaranteeing the clear display of MPEG-2 video sequence when conduct splicing of MPEG-2 system streams. we focus on the PES domain splicing considering video sequence. And we wish to make a base on the TS or PS domain splicing considering video sequence. For that, first, we compared and analyzed problems that is raised when different two PES streams are spliced and effects that affect the video sequence. And based on this analysis, we have searched for methods that resolve the cause of problems that can be happened in the display of video sequence directly in PES domain.

This paper presents the real-time PCB(Printed circuit board) inspection system that can detect thin open/short error using the line scan camera. After a overall introduction of our system, the outline of our inspection methods are described. The goal of our inspection system is the real time and detailed inspection using the line scan camera. To perform inspection processing in real-time, we utilize double buffering structure. In order to solve the problem of unexpectable pixels of PCB, we propose melting process which eliminates unexpectable pixels of PCB. The design and development of our prototype of PCB ins- pection system is discussed and test results are presented to show the effectiveness of the developed inspection algorithm.

In this paper, the proposed watermark algorithm is based on energy distribution of the subband coefficients in the frequency domain and edge of the original image in the spacial domain. Out of these information, the KeyMap which decides the embedded position of watermark is produced. And then the binary watermark is embedded into the wavelet coefficient of LL3 subband using KeyMap and LFSR(Linear Feedback Shift Register).

In this paper we present a barcode reader to decode two-dimensional symbology PDF-417 and propose a novel method to extract the bar-space patterns directly from the gray-level barcode image, which employs the location and the distance between extreme points of each row of the barcode image. This algorithm proves to be very robust from the high convolution distortion environments such as defocussing and warping, even under badly illuminating condition. If the scanned barcode image is a result of the convolution of a Gaussian-shaped point spread function with a bi-level image, popular image segmentation methods such as image thresholding can not distinguish between very narrow bar-space patterns with a couple of pixels wide. The proposed algorithm shows improved performance over current barcode readers.

This paper presents robust and efficient scene-based video watermarking method using visual rhythm (spatio-temporal slice) in compressed domain. Scene change can be detected easily using visual rhythm and video sequences are conveniently edited at the scene boundaries. Therefore, scene-based watermark embedding Process it a natural choice. Temporal spread spectrum can be achieved by applying spread spectrum methods to visual rhythm. Additive Gaussian noise, low-pass filtering, median filtering and histogram equalization attack are simulated for all frames. Frame sub-sampling is also simulated as a typical video attack Simulation results show that proposed algorithm is robust and efficient in the presence of such kind of attacks.

In this paper, we propose a watermarking technique that embeds a digital watermark into digital images for the proof of owner or copyright protection. Proposed algorithm is based on discrete wavelet transform. Zerotree structure defined by Shapiro's embedded zerotree wavelet(EZW) algorithm is used. In the proposed algorithm, a digital watermark is embedded on only significant wavelet coefficients chosen by QSWT for the robustness of the algorithm. In other words, only the values of significant wavelet coefficients are modified in accordance with the given watermark pattern. We use the relationship among neighboring coefficients when modifying chosen coefficients to keep good image quality. Visual recognizable patterns such as binary images are used as a watermark. The experimental results show that the proposed algorithm has robustness under a variety of attacks such as JPEG compression, sharpening and blurring and also show that it has a better performance in PSNR comparing with other algorithms.

This paper proposes a unique data embedding algorithm for the video sequence. It describes two processings: shot boundary detection and robust data embedding. First, for the shot boundary detection, instead of using absolute frame differences, block by block based relative frame differences are employed. Frame adaptive thresholding values are also employed for the better detection. Second, for the robust data embedding, we generate message template and then convolve and correlate it with carrier signal. And then we embed data on the time domain video sequence. By using these two methods, watermarks into randomly selected frames of shots. Watermarks are detected well even if several certain shots are damaged because we embed watermark into each shot equally.

This paper proposes a new watermarking algorithm to embed data in 1-bit images using edges for protecting illegal copies or modifications. This method is based on edge detection to decide the most invisible region. This paper also introduces a new shuffling method which embeds data in each blocks uniformly and finds Positions under the limited attacks. In experiment, we compare the proposed shuffling method with M. Wu's method.

In this paper, we propose the efficient digital watermarking scheme to transmit effectively the compressed medical image that embedded with watermarking data in mobile Internet access channel. The wireless channel error based on multiple access interference (MAI) is closely related to the length of spreading sequence in CDMA system. Also, the fixed length coded medical image with watermark bit stream can be classified by significance of source image. In the simulation, we compare the peak signal to noise ratio (PSNR) performance when the watermarked image with a simple symbol and when the watermarked image with a text file is transmitted using variable length of spreading sequences in case of limited length of spread sequence.

Windows provides Directshow for efficient multimedia streaming processings such as multimedia capture, storage, display and etc. Presently, many motion picture codecs and audio codecs are made to be used in Directshow framework and Windows also supports many codecs (MPEG4, H,263, WMV, WMA, ASF, etc.) in addition to a lot of useful tools for multimedia streaming processing. Therefore, Directshow can be effectively utilized for developing windows-based multimedia streaming applications such as visual monitoring systems which needs to store real-time video data for later retrieval. In this paper, we present our efforts for developing a Directshow Filter System supporting storage of motion pictures in various motion picture codecs. Our Directshow Filter system also provides an additional functionality of motion detection.

In this paper, we propose a new text watermarking method using space coding and PN sequence. A PN sequence generated from user message modifies the space between words in each line. The detection can be done without original text image using the average space with in the text. Experimental results show that proposed method has the invisible property and robustness to the attack such as the elimination of words in the text.

This paper presents an adaptive digital image watermarking scheme that uses successive subband quantization (SSQ) and perceptual modeling. Our approach performs a multiwavelet transform to determine the local image properties optimal and the watermark embedding location. The multiwavelet used in this paper is the DGHM multiwavelet with approximation order 2 to reduce artifacts in the reconstructed image. A watermark is embedded into the perceptually significant coefficients (PSC) of the image in each subband. The PSCs in high frequency subbands are selected by setting the thresholds to one half of the largest coefficient in each subband. After the PSCs in each subband are selected, a perceptual model is combined with a stochastic approach based on the noise visibility function to produce the final watermark.

This paper propose a rectification technique by applying the Projection matrices derived from perspective projection matrices estimated from self-calibrated stereo image pairs. The derivation is made such that two epipolar lines are in parallel. Rectified images are generated by reprojecting corresponding image points. For the performance analysis of this technique, vertical coordinates of rectified points are compare to those obtained by the technique[3].

In this paper, we propose a 3-D animation system to track and analyze motion of the human object. The proposed system consists of two separate layers: motion analysis layer and 3-D model registration layer. Following the MPEG-4 SNHC standard, we generate object motion using body definition and animation parameters. In the implemented system, we acquire human motion data from a single camera and extract body definition parameters from arbitrary VRML human models.

In this paper, we propose a method to extract vertical stripes projected on human face using Gabor filter, Previous work cannot extract continuous vertical stripes in the eye and mouth region due to their horizontal lines, Proposed method use Gator filter adaptively according to main frequencies and directions of stripes in each block. Experimental results show that Proposed method can extract continuous vertical stripes in the eye and mouth region

This paper presents a new variational framework for detecting and tracking moving objects in image sequence. Motion detection is performed using Level Set Model. The original frame is used to provide th moving object boundaries Then, the detection and the tracking problem are addressed in a common framework that employs a inward-outward curve evolution function. This function is minimized using a gradient decent method.

This paper describes a new feature based tracking system that can track moving objects with a pan-tilt camera. We extract corner features of the scene and tracks the features using filtering, The global motion energy caused by camera movement is eliminated by finding the maximal matching position between consecutive frames using Pyramidal template matching. The region of moving object is segmented by clustering the motion trajectories and command the pan-tilt controller to follow the object such that the object will always lie at the center of the camera. The proposed system has demonstrated good performance for several video sequences.

A study on image searching and management techniques is actively developed by user requirements for multimedia information that are existing as images, audios, texts data from various information processing devices. We had been studied an automatical shape region segmentation method using color. distribution and edge characteristics of moving images for. contents-base description. The Proposed method uses a color information quantized on human visual system and extracts overlapped regions to be matched by using edge characteristics of the image frame. The performance of the proposed method is represented by similarity for comparison to a segmented image and original image.

Automated video segmentation is important as the first step of video indexing, video retrieval and other uses. Unlike abrupt changes that are relatively easy to detect, gradual transitions like dissolve, fade-in and fade-out are rather difficult to detect. In this paper, we propose a method for detecting gradual transitions based on local statistics and less dependent to a given threshold level. Experimental results show that the proposed method detected about 85% of gradual transitions.

The image labeling and classification is one of the important tasks for a content-based image retrieval and an image understanding. This paper propose a new technique to label and classify natural images with a color-spatial histogram and production rules. We show that our proposed method is very efficient for a natural image composed of a few regions.

Scene change detection is the fundamental process of automatic video indexing and retrieving. In this paper we propose a method which utilizes both compressed and uncompressed domain methods to detect scene change in a video. Candidate locations for scene change are estimated from DC images and motion vector information in compressed domain. And candidate frames are verified using edge histogram distance and color histogram distance, in uncompressed domain. The experimental results show that scene change can be detected fast and correctly by proposed method.

We propose a new method for rotational image retrieval that it is based on highly related property between a spatial image and wavelet transform. The characteristics have an important role in the design of our algorithm. Our proposed algorithm for rotational image retrieval is to obtain same image or rotated image. Because our algorithm used an rotational image retrieval.

In this paper, we implement a 2.4 kbps STC speech codec using the TMS320C6201 DSP The main job for this work is twofold: one is to convert floating-point operation in the codec into fixed-point operation while maintaining the high resolution, and the other is to optimize the program to make it run in real time with memory size as small as possible. The implemented decoder uses 54.8 kbyte of program memory, 29.7 kbyte of data ROM and 55.2 kbyte of data RAM, respectively. It also uses about 45% of maximum computation capacity of TMS320C6201.

This paper presents an implementation method of phasor measurement device, which is based on the FPGA implementation of the sliding-DFT The design is verified by the timing simulation of its operation. The error effect of coefficient approximation and frequency deviation in the recursive implementation of the sliding-DFT is analytically derived and verified with the computer simulations.

본 논문에서는 디지털 오디오 증폭기의 필수적인 부분인, PCM신호를 PW신호로 변환하는 기법에 대하여 고찰한다. 비교적 낮은 해상도의 PWM 신호로 변환할 때 발생하는 문제점들을 살펴보고, oversampling과 통과 대역에서 잡음을 억제하는 noise shaping을 적용한 기법에 대하여 살펴본다. 본 논문에서는 디지털 오디오 증폭기에 적합한 몇 가지 noise shaping 필터를 소개하고, 그들을 사용하였을 때 오디오 품질을 평가하였다.

In this paper, we propose a wavelet based adaptive algorithm which improves the convergence speed and reduces computational complexity using the fast running FIR filtering efficiently We compared the performance of the proposed algorithm with time and frequence domain adaptive algorithm using computer simulation of adaptive noise canceler based on synthesis speech. As the result, the proposed algorithm is suitable for adaptive signal processing area using speech or acoustic field.

In this paper, we present an area-efficient programmable FIR digital filter using canonic signed-digit(CSD) coefficients, in which the number of effective nonzero bits of each filter coefficient is reduced by sharing the shift and add logics for common nonzero bits between adjacent coefficients. Also, unused shift and add logics for a low- magnitude coefficient are reassigned to an appropriate high - amplitude coefficient. In consequence, the proposed architecture reduces the hardware area of a programmable FIR filter by about 24% and improves performance about 6-7dB compared to other multiplierless FIR filters with powers-of-two coefficients.

본 논문에서는 지금까지 소개된 44.1KHz compact disc (CD)에서 48KHz digital audio tape (DAT)로의 표본화율 변환기법들에 대해서 가청 주파수 대역에서 100dB 이상의 dynamic range와 ±5x10­4dB 이하의 리플 크기를 유지할 수 있도록 각 기법들을 재설계하였으며, 메모리 요구량 및 계산량에 대해서 살펴보고자한다.

The final purpose of this paper is the enhancement of speech recognition rate under the matched telephone environment between training data and test data. To analyze the effect by the distortion of the changing telephone channel on every call, MFCC is used as the feature parameter and CMN, RTCN, and RASTA are used as channel compensation techniques. For each case, the variation of feature parameters of all phones is analyzed. And, we find recognition rates according to each compensation method using the continuous HMM recognizer, and examine the relationship between variation and recognition rate.

We propose an efficient mixture Gaussian synthesis method for decision tree based state tying that produces better context-dependent models in a short period of training time. This method makes it possible to handle mixture Gaussian HMMs in decision tree based state tying algorithm, and provides higher recognition performance compared to the conventional HMM training procedure using decision tree based state tying on single Gaussian GMMs. This method also reduces the steps of HMM training procedure. We applied this method to training of PBS, and we expect to achieve a little point improvement in phoneme accuarcy and reduction in training time.

This paper introduces the near field shading beamformer using widely known Chebyshev and Hanning window in the field of digital signal processing. The proposed shading beamformer improves the estimation of range as well as azimuth angle of targe residing in near field. A series of sensor weighting values are calculated from the FFT operation of given shading functions in time domain. This paper verifies the performance of the focused beamformer having the proposed shading sensor weights which are used to detect the range of target. Throughout computer simulations this paper exploits the performance improvement of the proposed shading beamformer as varying the frequency band of the received radiated signal along the non-uniform array.

In this paper, we present a sentence speech recognizer for twenty questions game. The proposed approaches for speaker-independent sentence speech recognition can be divided into two steps. One is extraction of the number of syllables in eojeol for candidate reduction, and the other is knowledge based language model for sentence recognition. For twenty questions game, we implemented speech recognizer using 956 sentences and 1095 eojeols. The results obtained in our experiments were 87% sentence recognition rate and 90.15% eojeol recognition rate.

A novel structuring element for the morphological filter is proposed in order to detect a small target at a long distance. The modeling of the structuring element is based on the real data and implemented by parametric model approach. Several synthetic

The edges of an image hold much of the information in that image. The edges tell where objects are, their shape and size, and something about their texture. An edge is where the intensity of an image moves from a low value to a high value. We introduce the edge detection using the differential operator with Sobel operator and describe a nonparametric Wilcoxon test based on statistical hypothesis testing for the detection of edges. This paper proposes an efficient edge detection using Van der Waerden's statistic in original and noisy images. We use the threshold determined by specifying significance level a and an edge-height parameter. Comparison with our statistical test and Sobel operator shows that Van der Waerden method perform more effectively in both noisy and noise-free images.

In this paper, we propose a new blind image watermarking method in wavelet domain. It is necessary to find out watermark insertion location in blind watermark. We use horizontal and vertical difference of LL components to select watermark insertion location, because increment or decrement of successive components is rarely changed in LL band. A pseudo-random sequence is used as a watermark. Experimental results show that the proposed method is robust to various kinds of attacks such as JPEG lossy compression, averaging, median filtering, resizing, histogram equalization, and additive Gaussian noise.

영상 검색을 위한 기존의 칼라 히스토그램 방법은 영상의 형상 정보를 포함하고 있지 않다. 본 논문에서는 로컬 형상 정보인 에지 정보에 칼라 정보를 접목시켰다. 각각의 세 종류의 에지(수평 에지, 수직 에지, 비 방향성 에지) 주위의 픽셀들에 대한 칼라 분포를 구한 후 그 각각의 칼라 분포를 기초로 구한 두 영상간의 거리 정보를 이용하여 영상들간의 유사도를 구했다. 따라서 본 논문에서 제안한 유사도는 한 종류의 에지 타입 (비 방향성 에지)을 이용한 방법에 비해 매칭 오류를 줄일 수 있었다. 실험 결과, 제안된 방법이 기존의 칼라 히스토그램과 에지 히스토그램을 이용한 방법에 비해 향상된 성능을 보였다.

In this paper, we propose an algorithm to decode Data Matrix two dimensional barcode symbology. We employ hough transform and bilinear image warping to extract the barcode region from the image scanned using a CMOS digital camera. The location of barcode can be found by applying Hough transform. However, barcode image should be warped due to the nonlinearity of lens and the viewing angle of camera. In this paper, bilinear warping transform is adopted to wa게 and align the barcode region of the scanned image. Codeword can be detected from the aligned barcode region.

As control programs are becoming larger and more complex, the capacity of a microprocessor in a TV set needs to have a CPU with 32 bits or more. Also the embedded system such like a digital TV set uses RTOS (Real Time Operating System)[1]. LGE Co. has developed a digital TV system that displays HD (High definition) signal with the AR7TDMI CPU and the commercial RTOS[2]. This thesis made improvements on problems caused by porting embedded LiNUX in limited size memory (2 mega byte) instead of RTOS[3]. The advantages of changing to embedded LiNUX take away running OS royalty that needed to make TV sets[4].

This paper presents a way of a novel FPGA implementation of DCT. It shows how to limit the required bits on each DCT processing step, instead of implementing high-cost 64-bit floating-point arithmetic of IEEE Std 754-1985 on FPGA. ID-DCT implementation has been done which operates at 30 frame per second with 1920${\times}$1080 resolution.

The rapid expansion of the Internet in the past few years has rapidly increased the availability of digital data such as audio, images and videos to the public. Therefore, The need for copyright protect of digital data are increasing in the internet. In this paper, Color image(RGB model) is transformed into LUV model, it includes the characteristics of, human vision and then the U or V component is transformed into 3-level wavelet transform. we can insert watermark to several objects of an image separately The experimental results showed that the proposed watermarking algorithm was better than to other RGB watermarking algorithm.

The proliferation of digitized media(audio, image and video) is creating a processing need for copyright enforcement schemes that protect copyright ownership. we argue that a watermark must be placed in perceptually significant components of a signal if it is to be robust to signal distortions and malicious attack. In this paper, RGB coordinate image is transformed into LUV coordinate, it include the characteristics of, Human vision and then the UV component is transformed into NxN block DCT transform. we propose a technique for embedding the watermark of visually recognizable mark into the middle frequency domain of image.

Data embedding has recently become important for protecting authority. In this paper, we Propose a robust data embedding technique for images. Our techniques are based on the convolution between message image and a random phase carrier. We add extra bits with carrier image to improve precision of detecting rate, moreover, we use block by block based cyclic correlation for the compensation of distortion. In experiment, we show that the proposed a1gorithm is robust to Stirmark 3.1. attacks.

Currently many devices reproduce electronic images in a variety of ways. However, the colors that are reproduced are different from the original color due to the differences in the gamut between devices. In this paper, a gamut mapping method utilizing a simultaneous mapping function and a lightness rescaling is proposed. This method enhance the local-color characteristics and lightness contrast. The experimental result shows that the overall contrast and the colorfulness were increased.

Recently, Digital TV viewer have been replacing cathode ray tubes (CRT) with Plasma display panel(PDP). But the chromaticity of the primaries are dependent on RGB input signals. And the colorimetry of PDP changes with gray scale and has a poor performance in color reproduction. In this paper we propose the enhanced algorithm of color reproduction considering nonlinear gamut mapping algorithm. In order to test performance of this algorithm we use the sample colors. As a result of experiments, it was confirmed that the color difference of the digital PDP using the proposed algorithm was considerably reduced.

In this paper, we present a hard disk drive embedded digital satellite receiver with scene change detection for video indexing. This receiver can store, retrieve and classify the broadcast data by implementing an interface between the conventional digital satellite receiver and digital storage media. Using this system, user can obtain more information for efficient video retrieval.

This paper describes extraction procedure for the center coordinates of steam generator tubes of Youngkwang NPP #6, which are arrayed in triangular patterns. Steam generator tube images taken with wide field-of-view lens and low-light lamp mounted on a ccd camera tend to have low contrast, because steam generator is sealed and poorly illuminated. The extraction procedures consists of two steps. The first step is to process the region with superior contrast in entire image of steam generator tubes and to extract the center points. Using the extracted coordinates in the first step and the geometrical array characteristics of tubes lined up in regular triangle forms, the central points of the rest region with low contrast are estimated. The straight lines from center point of a tube to neighbour points in horizontal and 60, 120$^{\circ}$ degree directions are derived. The intersections of straight line In horizontal direction and slant line in regular triangle direction are selected as the center coordinates of steam generator tubes. The Chi-square interpolation method is used to determine the line's coefficients in horizontal and regular triangle direction.

In this paper, a new watermarking method using an iterative scheme is proposed. The proposed method consists of two parts: watermark generation and its insertion. In the watermark generation, random signals with normalized distribution N(0,1) are produced and modified using an iterative scheme. In the watermark insertion, the resulting watermark signals are inserted into selected transform coefficients with various scaling parameters in order to achieve the invisibility and robustness against illegal attacks. The simulation results show that the proposed method has good performance for various attacks including image compression, filtering, etc.

The YCbCr color is widely used for digital video. In this paper, six different histogram algorithms are defined; the corresponding test images are then produced; and these images are then evaluated by observers using psychophysical techniques such as pair comparison; and show observer's preference data based on the psychophysical experiment.

Sharpness enhancement, which strengthen the edge(high frequency) of image, is widely studied for image processing research area. In this paper, psychophysical experiment is conducted by the 20 observers with simple linear unsharp masking for sharpness enhancement. The experimental results extracted using z-score analysis and linear regression suggests observer preferable sharpness enhancement method for digital television.

This Paper of use vision system, there is no DGPS's information, DGPS data value receives real time exactly without being influenced in surroundings environment because using vision system that is used in self-regulation traveling by car system. Therefore, conversion and DGPS of received in camera coordinate changing coordinate error correct and wish to grasp correctly position of vehicles.

Most of the Automatic Fingerprint Identification Systems define the ridge endings and bifurcation points as the minutia for matching. Therefore, the precise extraction of the minutia is critical in raising the efficiency and reliability of the system. The fingerprint images produced through the preprocessing may have the false minutia happened over the process and they can be the factors to decrease the system efficiency This paper suggests the algorithm, which removes the false minutia after extracting the candidate minutia from the thinned binary images of fingerprint images.

The block matching algorithm motion estimation is a soft-core for hardwired motion estimation block in MPEG-2, H.261 encoder. This motion estimation has been tested and verified to be valid for implementation of FPGA. Efficiency performance of the synthesized motion estimation was up to 89%, and the average PSNR between the original image and the motion-compensated image is 38dB.

A simple computer vision technology to measure the middle-ranged depth with mono camera and plain mirror is proposed The proposed system is structured wiか the rotating mirror in front of the fixed mono camera In contrast to the previous stereo vision system in which the disparity of the closer object is larger than that of the distant object, the pixel movement caused by the rotating mirror is bigger for the pixels of the distant object in the proposed system Being inspired by such feature in the proposed system the principle of the depth measurement based on the relation of the pixel movement and the distance of object have been investigated. Also, the factors to influence the precision of the measurement are analysed The benefits of the proposed system are low price and less chance of occlusion. The robustness for practical usage is an additional benefit of the proposed vision system.

This paper presents a fast block-matching algorithm to improve the conventional Three-Step Search (TSS) based method. The proposed Comparison Fast Block Matching Algorithm (CFBMA) begins with DAB for adaptive search range to choose searching method, and searches a part of search window that has high possibility of motion vector like other partial search algorithms. The CFBMA also considers the opposite direction to reduce local minimum, which is ignored in almost conventional based partial search algorithms. CFBMA uses the summation half-stop technique to reduce the computational load. Experimental results show that the proposed algorithm achieves the high computational complexity compression effect and very close or better image quality compared with TSS, SES, NTSS based partial search algorithms.

Digital video camera can be a useful tool to capture images for use in colorimeter. However, the RGB signals generated by different digital video camera are not equal for the same scene. The digital video camera for use in colorimeter is characterized based on the CIE standard colorimetric observer. One method of deriving a colorimetric characterization matrix between camera RGB output signals and CIE XYZ tristimulus values is Polynomial modeling. In this paper, 3${\times}$3 linear matrix and 3${\times}$l1 polynomial matrix is used to investigate the characterization performance of the professional digital video camera. In experimental results, it is demonstrated that proposed 3${\times}$3 linear matrix has a reasonable degree of accuracy for use in colorimeter.

인공위성영상을 이용하여 벡터 지도 생성은 지형에 따른 건물, 도로, 농지 등에 관한 벡터를 추출하는 작업이 필요하다. 이 작업의 정확도는 지도의 정확도와 상관관계가 있기 때문에 건물 추출의 정확성이 달라진다. 따라서 건물추출의 정확성을 향상시키기 위해 본 연구에서는 위성영상에서 건물 추출 알고리즘을 제안하였다. 이 알고리즘은 인공신경망을 이용하여 건물의 그림자를 추적하고 이를 중심으로 건물위치와 외형을 추정하는 알고리즘을 제안하고 실험하였으며, 양호한 결과를 얻었다.

Color reproductions in most ink jet printer are quite different from that of standard CRT (cathode ray tube) monitor display because of the nonlinear characteristic in subtractive color reproduction. Gray scale CCT(correlated color temperature) reproductions in a typical printer are vary with the input RGB level. A simple method for making constant gray scale CCT and gamma value in photo paper is proposed in this paper. The compensation of the CCT with white point of the photo paper under the CIE standard illuminant D65 and color correction has been confirmed using the LUT(look-up table) to compensate the CCT and gamma curve characteristic.

This paper has been studied a Image watermarking using PN coding operation where frequency band. By using wavelet transformation, This gets high frequency place HH2 where image watermark puts. Also this places that PN code and binary image are operated. And then, this paper has designed image watermarking index process and extract process, watermarked image which are to add noise is able to extract watermark.

본 논문은 워터마크의 일반 요건 중에서 가장 기술적으로 요구되는 강인성과 비인지성을 함께 발전시키기 위해서 구성되었다. 먼저, 주파수 영역을 이용한 워터마킹 기법 중에서 현재 압축 표준에서 사용되고, 지역적 인간 시각 시스템을 이용할 수 있기 때문에, 널리 사용되는 블록 단위의 DCT를 사용한다. 삽입 위치를 결정하기 위하여, 기존에 발표된 논문들의 강인성을 비교하여, 보다 나은 위치를 선정한다. 그리고, 인간 시각 시스템 중에서, 강한 에지의 인지성을 줄이기 위하여, 에지의 오리엔테이션을 보존하고 수평과 수직 성분보다는 45 도에 가까운 DCT 대각 성분에 워터마크의 강도를 높여준다 이는 워터마크 이후의 강한 에지 성분의 오리엔테이션 오차를 줄여주며, 에지의 비인지성을 향상시키는 결과를 가져온다.

In this paper, we present a digital authentication technique using content-based watermarking in digital images. To digest the image contents, Hopfield network is employed on the block-based edge image. The Hopfield function extracts the same tit fur similarly looking blocks so that the values are unlikely to change to the innocuous manipulations while being changed far malicious manipulations. By inputting the extracted bit sequence with secret key to the cryptographic hash function, we generate a watermark for each block by seeding a pseudo random number generator with a hash output Therefore, the proposed authentication technique can distinguish between malicious attacks and innocuous attacks. Watermark embedding is based on the block-based spread spectrum method in DCT domain and the strength of watermark is adjusted according to the local statistics of DCT coefficients in a zig-zag scan line in AC subband. The numerical experiments show that the proposed technique is very efficient in the performance of robust authentication.

This paper has been studied how to deal with moving objects in images when we mosaic them. The global motion between two images is biased due to the local motion from these moving objects, so it is very important how to eliminate the effects of them. In this paper contour-based approach for mosaicking images that contains moving objects is presented. Once we get the contours of images we can both eliminate the moving objects and mosaic the images. In this stage, hierarchical moving objects elimination technique is introduced. Experiment is done for Stefan tennis sequences to verify the proposed algorithm.

This paper proposes the simulation algorithm of rear lamp of car using a ray tracing method that models transmission and refraction of light frequently used in computer graphics At first. To reproduce an image accurately. incident on a sight of viewer, we propose the backward ray tracing method based on spectral distribution representing physical characteristics of illuminant and object used in real. We implement the reproduction algorithm of rear lamp image applying the Bouguer-Beer's law to an optical absorptive phenomenon. As the result, more realistic image can be reproduced.

In this paper, we present two types of vision algorithm that mobile robot has CCD camera. for obstacle avoidance. This is simple algorithm that compare with grey level from input images. Also, The mobile robot depend on image processing and move command from PC host. we has been studied self controlled mobile robot system with CCD camera. This system consists of digital signal processor, step motor, RF module and CCD camera. we used wireless RF module for movable command transmitting between robot and host PC. This robot go straight until recognize obstacle from input image that preprocessed by edge detection, converting, thresholding. And it could avoid the obstacle when recognize obstacle by line histogram intensity. Host PC measurement wave from various line histogram each 20 Pixel. This histogram Is ( x , y ) value of pixel. For example, first line histogram intensity wave from ( 0, 0 ) to ( 0, 197 ) and last wave from ( 280, 0 ) to ( 280, 197 ). So we find uniform wave region and nonuniform wave region. The period of uniform wave is obstacle region. we guess that algorithm is very useful about moving robot for obstacle avoidance.

This paper proposes some technical approaches for automatic detection of pulmonary nodules in chest X-ray images. We applied threshold technique for the lung field segmentation and extended the lung field by using morphological methods. A template matching technique was employed for automatic detecting nodules in lung area. Genetic algorithm(GA) was used in template matching(TM) to select a matched image from various reference patterns(simulated typical nodules). We eliminated the false-positive candidates by using histograms and contrasts. We used standard databases published by Japanese Society of Radiological Technology (JSRT) for correct results. Also we employ two-dimensional Gaussian distribution for some reference images because the shadow of lung nodules in radiogram generally shows the distributions. Nodules of about 89% were correctly detected by our scheme. The simulation results show that it is an effective method to indicate lesions on chest radiograms.

The feature points in the uncalibrated stereo vision should represent all the characteristics of an image in multiple resolution, have high precision, and have the robustness against mismatching. This paper proposed an algorithm which detects the corner points in multi-resolution for stereo computer vision. The algorithm has sub-pixel precision, rejects the mismatched points, and corrects the lens distortion. We show the performance of the algorithm by estimating the homography with it.

This paper presents an advanced auto focus camera system using the adaptive zoom tracking method. The proposed system can achieve an accurate zoom tracking with significantly reduced system memory.

This paper describes eye detection algorithm using edge information and Support Vector Machine (SVM). We adopt an edge detection and labelling algorithm to detect isolated components. Detected candidate eye pairs finally verified by SVM using Radial Basis Function (RBF) kernel. A detection rate over the test set has been achieved more than 90%, and compared with template matching method. this proposed method significantly reduced FAR.

DVR(Digital Video Recorder), a recently popular visual monitoring system, is required to support multi camera channel display(upto 16 channels) with various display modes in real-time. For such a multichannel display system, it is well known that tearing artifacts become more serious. Thus, one needs to design a display system for DVR so that it does not show tearing artifacts as much as possible, but keeps real-time display speed. In this paper, we present our efforts in designing and implementing a display system which rarely ever shows tearing artifacts, but without degradation of required real-time display speed, and which displays character information stably without blinking.

In this paper, we propose RTP(Real-Time Transport Protocol) based image transport system to transport still images in real-time after JPEG2000 compression, which is still image compression standard for next generation. To add RTP packet on UDP packet, the image transport system inserts packetizer and depacketizer process into transmitter and receiver of RTP data, respectively. We apply the proposed system to several image and compare the transport time to TCP-based method.

In this paper, we propose a new feature-based text watermarking for the binary text image. The structure of specific characters from preprocessed text image are modified to embed watermark. Watermark message are embedded and detected by the following method; Hole line disconnect using the connectivity of the character containing a hole, Center line shift using the hole area and Differential encoding using difference of flippable score points. Experimental results show that the proposed method is robust to rotation and scaling distortion.

In this paper, we present a low cost circuit design for a sentence speech recognition. The basic circuit of the designed sentence speech recognizer is composed of resistor, capacitance, OP Amp, counter and logic gates. Through a sentence recognition experiment, we can find the effectiveness of the designed sentence recognition circuit

In this paper, we present an improved TTS synthesis system using articulatory synthesis and concatenative synthesis. In concatenative synthesis, segments of speech are excised from spoken utterances and connected to form the desired speech signal. We adopt LPC as a parameter, VQ to reduce the memory capacity, and TD-PSOLA to solve the naturalness problem.

This paper represents a survey of a variety thresholding techniques including both global and local thresholding. Several thresholding methods are examined in detail to evaluate their performance based on a given set of test images. We also attempt to evaluate the performance of several thresholding methods for construction field documents image recognition system using a broken line structures, broken symbols and text, blurring of lines, symbols and text, noise in homogeneous areas measure as a criterion functions.

In this paper, speaker and context independent emotion recognition using speech signal is studied. For this purpose, a corpus of emotional speech data recorded and classified according to the emotion using the subjective evaluation were used to make statical feature vectors such as average, standard deviation and maximum value of pitch and energy and to evaluate the performance of the conventional pattern matching algorithms. The vector quantization based emotion recognition system is proposed for speaker and context independent emotion recognition. Experimental results showed that vector quantization based emotion recognizer using MFCC parameters showed better performance than that using the Pitch and energy Parameters.

In speech signal processing, it Is very important to detect the pitch exactly in speech recognition, synthesis and analysis. but, it is very difficult to pitch detection from speech signal because of formant and transition amplitude affect. therefore, in this paper, we proposed a pitch detection using the spectrum flattening techniques. Spectrum flattening is to eliminate the formant and transition amplitude affect. In time domain, positive center clipping is process in order to emphasize pitch period with a glottal component of removed vocal tract characteristic. And rough formant envelope is computed through peak-fitting spectrum of original speech signal in frequency domain. As a results, well get the flattened harmonics waveform with the algebra difference between spectrum of original speech signal and smoothed formant envelope. After all, we obtain residual signal which is removed vocal tract element The performance was compared with LPC and Cepstrum, ACF 0wing to this algorithm, we have obtained the pitch information improved the accuracy of pitch detection and gross error rate is reduced in voice speech region and in transition region of changing the phoneme.

This paper proposes a new GSC (Generalized Sidelobe Canceller) structure, called HLMS-GSC. Compared to Griffiths and Jim's GSC structure, the number of complex multiplication required is reduced to one half. The simulation results show that the minimum mean square errors and performance of nulling jammers by using HLMS-GSC are almost the same compared to Griffiths and Jim's GSC, although the complexity is reduced significantly. As a result, the proposed adaptive beamformer is good for real time implementation, since it has low complexity compared to previous GSC structures.

This paper propose a new GSC (Generalized Sidelobe Canceller) structure, called HFLMS-GSC. The number of complex multiplication required is reduced to one half compared to FLMS-GSC. The simulation results show that mean square error converging and jamming signal removing characteristics are almost the same compared to FLMS-GSC, although the complexity is reduced significantly. As a result, the proposed structure is good for real time implementation, since it has low complexity compared to previous GSC structures.

In this paper, we present a new method for generating a set of complementary sequences that can be simultaneously transmitted and yet compressed into a short pulse on reveive, and its application to ultrasound imaging. This new complementary sequences can be designed based on a filter bank theory. The new complementary sequences can be used to improve the SNR of ultrasound imaging without sacrificing the spatial resolution and frame rate, compared to conventional pulse-echo imaging. Computer simulations are performed to verify the proposed method.

This paper has been studied speech enhancement method in noisy environment. By mean of that we prefer human auditory mechanism which is perfect system and applied wavelet transform. Multi-resolution of wavelet transform make possible multiband spectrum analysis like human ears. This method was verified very effective way in noisy speech enhancement.

In general, speech coding methods are classified into the following three categories: the waveform coding, the source coding and the hybrid coding. In this paper, First, the reference waveform is detected after searching the pitch period by NAMDF similarity and similarity between the reference waveform and the waveform each pitch period. It made a decision whether the waveform is compressed with the threshold of similarity. If the waveform is compressed only magnitude and pitch information is transmitted into the input of G.723.1 vocoder. Performing through the G.723.1 vocoder, the waveform is restored with the magnitude and pitch information by PSOLA synthesis method. The result of simulation with proposed algorithm has a 31% reduction of bit rate than the standard 5.3kbps G.723.1 ACELP vocoder.

LSP(Line Spectrum Pairs) Parameter is used for speech analysis in vocoders or recognizers since it has advantages of constant spectrum sensitivity. low spectrum distortion and easy linear interpolation. However the method of transforming LPC(Linear Predictive Coding) into LSP is so complex that it takes much time to compute. Among conventional methods, the real root method is considerably simpler than others, but nevertheless, it still suffers from its jndeterministic computation time because the root searching is processed sequentially in frequency region. We suggest a method of reducing the LSP transformation time using voice characteristics The proposed method is to apply search order and interval differently according to the distribution of LSP parameters. in comparison with the conventional real root method, the proposed method results in about 46.5% reduction. And, the total computation time is reduce to about 5% in the G.723.1 vocoder.

There is a growing interest in speaker verification, which verifies someone by his/her voices. This paper explains the traditional text-dependent speaker verification algorithms, DTW and HMM. This paper also introduces SVM and how this can be applied to speaker verification system. Experiments were conducted with Korean database using these algorithms. The results of experiments indicated SVM is superior to other algorithms. The EER of SVM is only 0.5% while that of HMM is 5.4%.

In this paper, a speech enhancement system using microphone array with adaptive Post-Processing is implemented in real-lime with TMS320C6201 DSP. It consists of delay-and-sum beamformer and adaptive post-processing filters with NLMS (Normalized Least Mean Square) algorithm. THS1206 ADC is used for collection of 4-channel microphone signals. Sizes of program memory, data ROM and data RAM of the implemented system are 15,744, 748 and 47,540 bytes, respectively. Finally 21.839${\times}$106 clocks per second is required for real-time operation.

In this paper, we propose a VLSI architecture for computation of the SVM(Support Vector Machine) that has become established as a powerful technique for solving a variety of classification, regression, and so on. When we compare the proposed systolic arrays with the conventional method, our architecture exhibits a lot of advantages in terms of latency and throughput rate.