公开(公告)号：KR1020050063908A

公开(公告)日：2005-06-29

申请号：KR1020030095153

申请日：2003-12-23

Applicant: 한국과학기술원

Inventor： 박현욱 ,  박영서

IPC: G06T3/40 ,  H04N1/393

Abstract: 본 발명은 영상의 크기 변환 방법에 관한 것으로서, 더욱 상세하게는 압축된 형태로 저장된 정지영상이나 동영상을 압축된 환경에서 영상의 크기를 변환하는 방법에 관한 것이다.
본 발명에 따른 블록 이산코사인변환 도메인에서의 임의의 영상 크기 변환 방법은, n (양의 정수)개의 입력 블록(block)의 DCT계수에 대해 각각 역 DCT를 수행하여 n개의 공간 도메인 신호를 얻는 역 DCT 단계; 상기 n개의 공간 도메인 신호를 하나의 신호로 연결하여 1 블록을 만들고, 당해 1 블록에 대해 DCT를 수행하여 DCT 계수를 얻는 DCT 단계; 및 상기 DCT계수의 고주파수 쪽의 계수를 버리고 저주파수 쪽의 계수를 취하여 n-폴드 다운샘플링(downsampling)을 구현하는 다운샘플링 단계;를 포함하여 이루어진다.

公开(公告)号：KR1020030096504A

公开(公告)日：2003-12-31

申请号：KR1020020033019

申请日：2002-06-12

Applicant: 한국과학기술원

Inventor： 이철수 ,  박현욱

IPC: H04N1/41

CPC classification number: H04N19/13 ,  H04N19/91

Abstract: PURPOSE: A method and an apparatus for compressing multiple code data using binary arithmetic coding are provided to improve compression and release speed by using binary arithmetic coding, and increase a compression rate by coding bits with conditional probability to binary code. CONSTITUTION: Multiple code data is transformed into binary type data(S110). An MSB(most significant bit) is coded by using probability having a null condition(S120). A second bit is coded by using conditional probability having considered a value of the MSB(S130). A kth bit is coded by using conditional probability having considered all values of a superior position(S140).

Abstract translation: 目的：提供使用二进制算术编码压缩多个代码数据的方法和装置，以通过使用二进制算术编码来提高压缩和释放速度，并且通过将具有条件概率的比特编码为二进制码来提高压缩率。 构成：将多个代码数据转换为二进制数据（S110）。 通过使用空值条件的概率对MSB（最高有效位）进行编码（S120）。 通过使用已经考虑到MSB的值的条件概率对第二比特进行编码（S130）。 通过使用考虑了上位的所有值的条件概率来编码第k位（S140）。

公开(公告)号：KR100405819B1

公开(公告)日：2003-11-14

申请号：KR1020010002096

申请日：2001-01-15

Applicant: 한국과학기술원

Inventor： 이철수 ,  박현욱

IPC: H04N1/41

CPC classification number: H04N19/90 ,  H04N19/13 ,  H04N19/91

Abstract: A method for compressing and restoring data of a binary image is provided to bind four pixels into one cell for compression, without compressing the image by the pixel, and to compress each cell twice, so as to enhance a compression ratio of the binary image. In a first step to compress S-value, a preprocessing unit(301) performs preprocessing with a rearranged image, to extract S-value of each cell(302). A delay(302) delays the extracted S-value. A probability estimator(304) for S-value estimates the probability for the S-value, and estimates the probability for a current symbol. A first entropy encoder(305) for S-value performs entropy encoding for the estimated probability using arithmetic coding for outputting through a multiplexer(312). In a second step to compress C-value of each cell using the compressed S-value, the image passing the preprocessing unit is stored in an image buffer(306). S-value of a cell to be currently compressed is calculated(307). C-value is extracted from the image buffer(309). A delay(310) delays the C-value. A probability estimator(308) for C-value estimates the probability for the C-value, and estimates the probability for a current symbol. And a second entropy encoder(311) for C-value performs entropy coding for the estimated probability using arithmetic coding for outputting through the multiplexer.

Abstract translation: 提供了一种用于压缩和恢复二值图像的数据的方法，用于将四个像素压缩成一个压缩单元，而不用像素压缩图像，并将每个单元压缩两次，以提高二值图像的压缩比。 在压缩S值的第一步骤中，预处理单元（301）利用重新布置的图像执行预处理，以提取每个单元的S值（302）。 延迟（302）延迟提取的S值。 用于S值的概率估计器（304）估计S值的概率，并估计当前符号的概率。 用于S值的第一熵编码器（305）使用用于通过多路复用器（312）输出的算术编码对所估计的概率执行熵编码。 在使用压缩的S值压缩每个单元的C值的第二步骤中，经过预处理单元的图像被存储在图像缓冲器（306）中。 计算当前要压缩的单元的S值（307）。 从图像缓冲器中提取C值（309）。 延迟（310）延迟C值。 C值的概率估计器（308）估计C值的概率，并估计当前符号的概率。 并且用于C值的第二熵编码器（311）使用算术编码对所估计的概率进行熵编码以通过多路复用器输出。

公开(公告)号：KR100389807B1

公开(公告)日：2003-07-02

申请号：KR1020010030007

申请日：2001-05-30

Applicant: 한국과학기술원

Inventor： 박근형 ,  이철수 ,  박현욱

IPC: H04N19/167

CPC classification number: H04N19/17 ,  H04N19/10 ,  H04N19/63

Abstract: A region of interest coding method based on the SPIHT algorithm, in which more bits are allocated with respect to a region of interest to quickly compress and restore an image, is provided. The region of interest coding method is applied to a multiple resolution coding operation. In the coding method, an image is coded using the SPIHT algorithm before a region of interest is set. When a region of interest is defined based on a region of interest mask and a parent mask, coefficients are coded with respect to a bit plane defined by a user, and a compensation coding operation is performed with respect to coefficients not related with the region of interest. The coding operation is performed based on the SPIHT algorithm by the amount of bits permitted in the bit plane of the remaining coefficients. Therefore, it is possible to implement compression and restoration of various images.

Abstract translation: 提供了基于SPIHT算法的感兴趣区域编码方法，其中关于感兴趣区域分配更多比特以快速压缩和恢复图像。 感兴趣区域编码方法被应用于多分辨率编码操作。 在编码方法中，在设置感兴趣区域之前，使用SPIHT算法对图像进行编码。 当基于感兴趣区域掩码和父掩码来定义感兴趣区域时，相对于由用户定义的位平面来对系数进行编码，并且对与不与该区域相关的系数执行补偿编码操作 利益。 编码操作基于SPIHT算法通过剩余系数的位平面中允许的位数来执行。 因此，可以实现各种图像的压缩和恢复。

公开(公告)号：KR1020020019854A

公开(公告)日：2002-03-13

申请号：KR1020000053215

申请日：2000-09-07

Applicant: 한국과학기술원

Inventor： 전재호 ,  박영서 ,  박현욱

IPC: H03M7/40

CPC classification number: H04N19/13 ,  H03M7/40 ,  H04N19/91

Abstract: PURPOSE: A high-speed variable length decoder using plane partition is provided to improve a total speed of a variable length decoder by performing simultaneously a parallel matching process of decoding data and an input process of a decoding code length signal. CONSTITUTION: Two input planes(11) are used for receiving data. The input planes(11) are separated to each other. A programmable logic array(5,5') is connected with a one input plane of the separated input planes(11). A subtracter(19) and an adder(20) are connected with an output terminal of the programmable logic array(5,5'). Code length signals of decoded information of a programmable logic array(5,5') are inputted into each input plane(11) and the subtracter(19), respectively. The remaining bit stream is checked by using the code length signal provided from the programmable logic array(5,5'). A supply control operation of a decoding bit stream is performed by using the code length signal provided from the programmable logic array(5,5').

Abstract translation: 目的：提供使用平面分区的高速可变长度解码器，通过同时进行解码数据的并行匹配处理和解码码长信号的输入处理来提高可变长度解码器的总速度。 构成：两个输入平面（11）用于接收数据。 输入平面（11）彼此分离。 可编程逻辑阵列（5,5'）与分离的输入平面（11）的一个输入平面连接。 减法器（19）和加法器（20）与可编程逻辑阵列（5,5'）的输出端子连接。 可编程逻辑阵列（5,5'）的解码信息的码长信号分别输入到每个输入平面（11）和减法器（19）中。 通过使用从可编程逻辑阵列（5,5'）提供的码长信号来检查剩余比特流。 通过使用从可编程逻辑阵列（5,5'）提供的码长信号来执行解码比特流的电源控制操作。

公开(公告)号：KR100301013B1

公开(公告)日：2001-09-06

申请号：KR1019980031431

申请日：1998-08-01

Applicant: 삼성전자주식회사 ,  한국과학기술원

Inventor： 이영렬 ,  박현욱

IPC: H04N19/82

Abstract: 본 발명은 영상데이터의 인코딩 및 디코딩할 때 발생하는 양자화효과 감소를 위한 루프필터링 방법 및 장치에 관한 것으로서, 그 방법은 역양자화된 영상데이터의 역양자화 계수의 분포 및 이전 프레임과 현재 프레임의 차를 나타내는 움직임 벡터를 이용하여 영상데이터의 루프 필터링 필요성 여부를 나타내는 플래그를 추출하는 플래그추출단계; 및 추출된 플래그가 루프 필터링이 필요함을 나타내면, 플래그에 상응하는 영상데이터를 소정의 방법에 의해 필터링하는 필터링단계를 포함함을 특징으로 한다.
본 발명에 의하면, 플래그와 적응필터를 사용하므로써 양자화 효과를 감소시키고, 필터링의 계산량을 줄이는데 큰 도움이 된다. 그리고 하드웨어 복잡성 측면에서 곱셈과 나눗셈 없이 병렬처리에 의해 수행될 수 있다.

公开(公告)号：KR100269125B1

公开(公告)日：2000-10-16

申请号：KR1019970055067

申请日：1997-10-25

Applicant: 한국과학기술원 ,  삼성전자주식회사

Inventor： 이영렬 ,  박현욱

IPC: H04N19/85

CPC classification number: H04N19/86 ,  H04N19/117 ,  H04N19/136 ,  H04N19/139 ,  H04N19/14 ,  H04N19/176 ,  H04N19/18 ,  H04N19/182 ,  H04N19/20 ,  H04N19/44 ,  H04N19/46 ,  H04N19/60 ,  H04N19/61 ,  H04N19/80

Abstract: PURPOSE: A method and an apparatus for post processing of video data for reducing a quantization effect are provided to reduce a quantization effect by using a semaphore and an adaptive filter. CONSTITUTION: A semaphore extraction portion(100) extracts a semaphore by using an inversely quantization coefficient of inversely quantized video data and a difference between a previous VOP(Video Object Plane) and a current VOP. A deblocking filter(110) performs deblocking filtering for decoded video data by checking a blocking semaphore. An outlier compensation portion(120) detects and compensates a coner outlier of deblocking filtered data. A deringing filter(130) checks a ringing semaphore and performs a deringing filtering process for the corner outlier data.

Abstract translation: 目的：提供一种用于后处理用于降低量化效果的视频数据的方法和装置，以通过使用信号量和自适应滤波器来减少量化效应。 构成：信号量提取部分（100）通过使用逆量化视频数据的反量化系数和先前VOP（视频对象平面）与当前VOP之间的差异来提取信号量。 去块滤波器（110）通过检查阻塞信号来对解码的视频数据进行解块滤波。 异常值补偿部分（120）检测并补偿去块滤波数据的更高值。 分页过滤器（130）检查振铃信号并对拐角异常值数据执行去噪过滤处理。