• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.8B
• /
• pp.1506-1517
• /
• 1999

In this paper, we propose two fast block matching algorithm using the distribution of mean absolute difference (MAD) at the search region overlapped with neighbor blocks and pixel subsapmling. The proposed methods use the lower and upper bound of MAD at the overlapped search region which is calculated from the MAD of neighbor block at that search position and MAD between the current block and neighbor block. In the first algorithm, we can reduce the computational complexity by executing the block matching operation at the only necessary search points. That points are selected using the lower bound of MAD. In the second algorithm, we use the statictical distribution of actual MAD which exists between the lower bound and upper bound of MAD. By using the statistical distribution of actual MAD, we can significantly reduce the computational complexity for motion estimation. after striking space key 2 times.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.5
• /
• pp.48-55
• /
• 1999

본 논문에서는 움직임 추정시 인접된 블록간의 탐색 영역이 중첩됨을 이용하여 전역 탐색 기법과 동일한 성능을 유지하면서 계산량을 현저히 감소시킬수 있는 고속 움직임 추정 기법을 제안하였다. 제안한 기법에서는 현재 움직임 추정을 행하고자 하는 블록에 대한 탐색점 중에서 인접된 블록과 중첩되는 탐색점에 대하여는 먼저 평균 절대 오차 (mean absoulte difference; MAD)를 계산할 필요가 있을지에 대한 판별을 행한 뒤 MAD 계산이 필요한 경우에 대하여서만 MAD를 구한다. MAD 계산 여부에 대한 판별에는 현재 블록과 인접 블록간의 MAD 와 인접 블록의 각 탐색점에 대한 MAD를 이용한다. 여기에 사용된 현재 블록과 인접 블록간의 MAD는 각 블록에 대하여 한번만 계산하면 되고, 인접 블록의 각 탐색점에 대한 MAD는 이미 구해져 있기 때문에 한번의 MAD 계산을 추가함으로써 탐색점 수를 현저히 감소시킬 수 있었다. 컴퓨터 모의 실험 결과로부터 제안한 방법이 전역 탐색 알고리즘과 동일한 성능을 유지하면서 많은 계산량의 감소를 얻을 수 있음을 확인 할 수 있었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.9
• /
• pp.1650-1657
• /
• 2007

In this paper, we propose an Overlapped Block and an Hanning Window to prevent a Block Discontinuity when we use an algorithm to eliminate ringing artifact which is based on a block structure. The algorithm to eliminate ringing artifact operates with a block structure and 24-RGB data and is based on a modified K-means algorithm. The proposed overlapped block method is piled up one on another per an half of the size of unit-block when an input image is split into several unit-blocks. Therefore, we define a data unit as the unit-block of the block size, $16{\times}16$ pixels. We reconstruct the processed data units into the original form of input image by using an isotropic form of Hanning Window. Finally, in order to evaluate the performance of the abovementioned algorithms, we compare three image, an input image with ringing artifact, an image result obtained by conventional method (non-overlapped), and an image result obtained the proposed method.

• Journal of Broadcast Engineering
• /
• v.24 no.5
• /
• pp.785-790
• /
• 2019

In this paper, a new bilateral frame rate up-conversion algorithm using adaptive overlapped block motion compensation is proposed. In this algorithm, the adaptive overlapped block motion compensation is based on the motion complexity of the reference region. As the motion complexity is determined by the size of the previously coded motion estimation prediction, the overlapped block motion compensation method is selected without any additional computational complexity. Experimental results show that the proposed algorithm provides better image quality than conventional methods both objectively and subjectively.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.36S no.12
• /
• pp.58-66
• /
• 1999

In this paper, an advanced histogram equalization algorithm for contrast enhancement is presented. Histogram equalization is the most popular algorithm. Global histogram equalization is simple and fast, but its contrast enhancement power is relatively low. Local histogram equalization, on the other hand, can enhance overall contrast more effectively, but the complexity of computation required is very high. In this paper, a low pass filter type mask is used to get a sub-block histogram equalization function to more simply produce the high contrast. The low pass filter type mask is realized by partially overlapped sub-block histogram equalization (POSHE). With the proposed method. the computation overhead is reduced by a factor of about one hundred compared to that of local histogram equalization while still achieving high contrast.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.11
• /
• pp.117-126
• /
• 2014

In this paper, we propose an efficient algorithm which can relocate the datablock of an aircraft when it is overlapped. If the datablock which represents the aircraft information in the control display is overlapped, relocation without overlapping is necessary because it is difficult to control the air traffic in this situation. The proposed algorithm relocates the data block with minimum movement by considering the characteristics of datablock. The moving distance of minimum movement is calculated using the height or width of rectangle which is created during overlapping. And the moving direction of minimum movement is calculated by considering the directivity of the datablock. When the distance between the target symbol and datablock is distant enough, the relocation is carried out using the existing algorithm as a special case. The proposed algorithm shows improved performance in comparison with the existing algorithm due to the fact that it considers many different cases of the datablock.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.155-157
• /
• 2020

본 논문은 딥러닝 기반의 중첩 블록 단위 이미지 압축 방법에 대하여 제안한다. 이미지를 블록 단위로 나누어 압축을 진행하며, 그 과정에서 생길 수 있는 블록화 현상을 제거하기 위해 블록의 주변부를 압축에 이용한다. 이로 인한 추가적인 부하를 제거하기 위해 복호화에 필요한 부하 영역을 제거한 뒤, 복호화기에서 특징맵들을 병합하여, 제거한 영역에 대해 주변 블록의 정보를 사용하여 복호화를 진행한다. 압축을 진행하고자 하는 이미지의 크기에 따라 급격하게 증가하는 요구 메모리를 이미지의 크기에 상관 없이 고정된 작은 메모리로 이미지 압축을 진행할 수 있다는 장점이 있다. 실험 결과로써 4K 이미지를 통해서 복원된 화질과 메모리의 사용량을 측정한 결과, 동일한 화질을 유지함과 동시에, 기존 방법대비 약 500배 적은 메모리 사용량을 보인다.

• Journal of Broadcast Engineering
• /
• v.9 no.1
• /
• pp.54-63
• /
• 2004

A new motion compensated frame Interpolation (MCI) algorithm by block based motion estimation (BME) is proposed. The block for the BME is composed of a large overlapped block for practical object motion estimation (ME) and a small block (which has a coinciding center with the ME-block) for the more precise motion compensated image description. Pixels in the block for the ME are sub-sampled to reduce computational complexity. The proposed method is executed with the various ME-blocks which have different size and sub-sampling ratio, and compared to the conventional method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.9A
• /
• pp.1401-1410
• /
• 1999

In this paper, we propose a fast block matching algorithm using the lower and upper bound of mean absolute difference (MAD) which is calculated at the search region overlapped with neighbor blocks. At first, we calculate the lower bound of MAD and reduce the search point by using this lower bound. In this method, we can get good prediction error performance close to full search block matching algorithm (FSBMA), but there exists some computational complexity that has to be reduced. Therefore, we further reduce the computational complexity by using pixel subsampling besides the lower and upper bound of MAD. Experimental results show that we can remarkably reduce the computational complexity with good prediction error performance close to FSBMA.

• Proceedings of the IEEK Conference
• /
• 2000.09a
• /
• pp.393-396
• /
• 2000

본 논문에서는 영상의 화소값으로부터 추출된 유사 특징점(quasi-feature point)을 이용한 이미지 모자이킹 알고리즘을 제안한다. 유사 특징점의 선택은 전역 정합(global matching)의 결과로부터 중첩된 영역을 4개의 부영역(sub-area)으로 분할하고, 각각의 분할된 부 영역에서 국부 분산(local variance)의 크기가 큰 블록을 선정, 이 블록의 중심 화소를 유사 특징점으로 선택한다. 유사 특징점에 대한 정합은 카메라 이동에 따른 왜곡(distortion)과 조명의 변화를 고려한 블록 정합 알고리즘(block matching algorithm)을 이용한다.