• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.2
• /
• pp.455-461
• /
• 2011

In object tracking using the block mating algorithm, it is not proper to use a fixed matching block to track an object of which size may be various and can be changed at any time. This paper defines an adaptive matching block for the dynamic environment and proposes a block matching algorithm for it. The matching block is composed of a main-block of $10{\times}10$ pixels and 8 sub-blocks of $6{\times}6$ pixels in a wide area of $42{\times}42$ pixels, the main-block located its center is used as an object block, and the sub-blocks located its boundary are used as candidates for the object block. The proposed algorithm extracts the object blocks from the sub-blocks by using their motion vectors for 10 previous frames and performs the block matching with the main block and them. The experiments for perform estimation show that the proposed algorithm extracts just valid object blocks from the matching block and keeps an object having free movement in image center area.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.04b
• /
• pp.879-882
• /
• 2001

장면 전환 검출을 위한 기존의 주요 방법은 대부분 화소값 차이, 통계적 차이, 그리고 일반적으로 가장 많이 사용하는 히스토그램 비교법등이었다. 이러한 방법에서는 각 비디오 프레임내에 포함되어 있는 영상의 내용 변화에 의한 장면 전환을 검출하기가 어렵고 히스토그램이 비슷한 경우 장면 전환 검출을 할 수 없었다. 본 논문에서는 비디오 시퀀스로부터 빠른 움직임을 포함하는 장면의 변화를 검출하는 객체의 중심점을 이용한 영역분할기법을 제안하였다. 제안하는 방법은 두 개의 프레임간 차영상을 이용해서 명암차를 추출한 후 화소값과 객체 영역의 중심점을 구한 후 사등분 분할하여 영역객체의 평균 및 분산값을 이용해서 내용에 의한 장면 전환도 검출하였다. 실험결과 제안된 방법에 의한 장면 전환 검출은 기존의 다른 방법에 비해 더 나은 성능을 얻었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.7
• /
• pp.1383-1388
• /
• 2016

Although image signal processing is used in many fields, degradation takes place in the process of transmitting image data by several causes. CWMF, A-TMF, and AWMF are the typical methods to eliminate noises from image data damaged under salt and pepper noise environment. However, those filters are not effective for noise rejection under highly dense noise environment. In this respect, the present study proposed an algorithm to remove in salt and pepper noise. In case the center pixel is determined to be non-noise, it is replaced with original pixel. In case the center pixel is noise, it segments local mask into 4 directions and uses linear interpolation to estimate original pixel. And then it applies spatial weight to the estimated pixel. The proposed algorithm shows a high PSNR of 24.56[dB] for House images that had been damaged of salt and pepper noise(P = 50%), compared to the existing CWMF, A-TMF and AWMF there were improvements by 16.46[dB], 12.28[dB], and 12.32[dB], respectively.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.12
• /
• pp.1844-1850
• /
• 2016

The bilateral filter can reduce the noise while preserving details computing the filtering output at each pixels as the average of neighboring pixels. In this paper, we propose a real-time system based on window division. Overall performance is increased due to the parallel architectures which computes five rows in the kernel window simultaneously but with pipelined scheduling. We consider the tradeoff between the filter performance and the hardware cost and the bit allocation has been determined by PSNR analysis. The proposed architecture is designed with verilogHDL and synthesized using Dongbu Hitek 110nm standard cell library. The proposed architecture shows 416Mpixels/s (397fps) of throughput at 416MHz of operating frequency with 132K gates.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.7
• /
• pp.989-995
• /
• 2022

Currently, the demand for image processing is increasing due to the development of IT technology, and active research is being conducted. Since image data generates image noise due to various external causes, and thus degrades the performance of the image, noise removal is essential. Salt and Pepper noise is a representative image noise, and various studies are being conducted to remove it. Existing algorithms include A-TMF, AFMF, LIWF, but these have the disadvantage that their performance is somewhat insufficient. Therefore, in this paper, we propose an algorithm that performs filtering using linear interpolation with effective pixels existing around the central pixel only in case of noise after performing noise judgment in order to efficiently remove salt and pepper noise. In order to judge the performance of the proposed algorithm, it was compared using the processed image of the previously studied algorithm and PSNR.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.422-424
• /
• 1999

본 논문에서는 질감 특징과 신경회로망을 이용한 유방암조직영상의 분할 방법을 제안한다. 신경회로망의 입력 노드에 사용될 질감 특징을 얻기 위해 10개의 영상에 대해 각 영역(양성세포핵, 음성세포핵, 배경)에서 10개씩의 화소를 선택하고, 그 화소를 중심으로 하는 5$\times$5 영역 30개를 획득, 총 300개의 영역에 대해 R, G, B 각각의 밴드에서 18개의 질감특징을 추출한다. 54개의 입력노드, 28개의 은닉노드, 3개의 출력노드의 구조를 가진 신경회로망을 구성하고, 역전파 학습 알고리즘을 사용하여 신경회로망을 최대오차율이 10-3보다 작을 때까지 학습시킨다. 학습에 의해 획득되어진 분류기를 이용하여 유방암 조직 세포영상을 양성세포핵, 음성세포핵, 배경부분으로 자동 분할한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2011.11a
• /
• pp.49-52
• /
• 2011

본 논문에서는 구간선형 모델을 적용하여 낮은 복잡도를 가지는 LSC(Lens-Shading Correction) 알고리즘을 제안한다. 제안한 알고리즘은 각 화소와 렌즈 중심점으로부터 거리를 정수형으로 계산하고, 이 정수를 거리에 대한 LSC 이득값이 저장된 LUT(Look-Up Table)에 대한 주소로 적용하여, 입력 화소 값에 곱함으로써 LSC를 수행한다. 거리를 구하려면 제곱근 회로가 추가되어야 한다. LUT에 저장된 이득값은 원점으로부터의 거리에 대한 평균 이득값을 저장하고 있기 때문에, 제곱근 계산에 높은 정밀도를 할애하여도 LSC 보상된 영상의 화질에 미치는 영향은 높지 않으므로 정수형 제곱근 연산을 수행한다. 제곱근 계산은 구간 선형화하여 단지 덧셈과 쉬프트 연산만으로 제곱근 연산을 완료할 수 있도록 간략화 하였다. 제안한 알고리즘을 양산 중인 일반 카메라 모듈에 적용한 결과, 카메라모듈 제조업체의 LSC 평가 기준을 상회하는 수준으로 나타나며, 구현될 하드웨어 복잡도가 매우 낮아서 모바일 카메라 구현에 매우 적합하다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.903-906
• /
• 2004

본 논문에서는 얼굴 표정 인식 시스템을 위한 얼굴 영역 추출 알고리즘을 제안한다. 이는 입력 영상으로부터 얼굴 후보 영역을 추출하고, 추출된 얼굴 후보 영역에서 눈의 위치를 정확히 추출한다. 그리고 추출된 눈 영역들의 정보와 타원 방정식을 이용하여 최종 얼굴 영역을 추출한다. 얼굴 후보 영역은 HSI 칼라 좌표계에 기반한 적응적 피부색 구간 범위를 설정하여 추출하였다. 추출된 얼굴 후보 영역에서의 눈 영역 추출을 위해 밝기 정보를 이용하여 먼저 눈의 후보 화소들을 추출하고, 레이블링 과정을 통하여 영역별로 그룹화하였다. 각 후보 영역들의 화소 수, 가로세로비 및 위치 정보를 고려하여 최종 눈 영역을 추출하였다. 추출된 두 눈 영역에서 무게중심을 구하고 이를 이용하여 장축과 단축을 설정하여 타원방정식을 이용 최종 얼굴 영역을 추출하였다. 제안된 알고리즘은 조명 변화, 다양한 배경들을 가지는 얼굴 영상에서도 정확히 얼굴 영역을 추출할 수 있었다.

• Korean Journal of Optics and Photonics
• /
• v.9 no.6
• /
• pp.362-367
• /
• 1998

Center Distance Finding (CDF) is a technique to find the sphere diameter by measuring the distance between two contacting spheres. The focal spots of the sphere clusters are formed in the back-focal plane by the transmission-mode optical microscope with the pseudothermal illumination source. Digital images taken by the CCD camera were processed by the software called Global Lab Image. The centers of the focal spots are found and the spot positions are expressed in terms of the CCD pixel elements, whose coordinate are calibrated by a heterodyne interferometer. The new CDF measurement system has been developed, which are more advantageous in time and convenience than the existing system, while the measurement uncertainly remains sufficient for its use as a magnification standard for optical microscopy. Two kinds of polystyrene spheres whose nominal diameters 3 and 10 $\mu\textrm{m}$ (NIST SRM 1962 and 1960) are measured with the uncertainly less than 1% at the confidence level of 99%, and the results are compared with the results of National Institute of Standards and Technology.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.2
• /
• pp.92-98
• /
• 2017

Generally, image contains geometrical and radiometric errors. Census transform can solve the stereo mismatching problem caused by the radiometric distortion. Since the general census transform compares center of window pixel value with neighbor pixel value, it is hard to obtain an accurate matching result when the difference of pixel value is not large. To solve that problem, we propose a census transform method that applies different 4-step weight for each pixel value difference by applying an assistance window inside the window kernel. If the current pixel value is larger than the average of assistance window pixel value, a high weight value is given. Otherwise, a low weight value is assigned to perform a differential census transform. After generating an initial disparity map using a weighted census transform and input images, the gradient information is additionally used to model a cost function for generating a final disparity map. In order to find an optimal cost value, we use guided filtering. Since the filtering is performed using the input image and the disparity image, the object boundary region can be preserved. From the experimental results, we confirm that the performance of the proposed stereo matching method is improved compare to the conventional method.