专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

51. 发明公开

KR1020040002115A 역 이산 코사인 변환 장치 无效
标题翻译：反向离散COSINE变换（IDCT）设备
公开(公告)号：KR1020040002115A
公开(公告)日：2004-01-07
申请号：KR1020020037545
申请日：2002-06-29
申请人： 매그나칩 반도체 유한회사
发明人： 김상철
IPC分类号： H04N19/42 , H04N19/625
CPC分类号： H04N19/42 , H04N19/625
摘要： PURPOSE: An inverse discrete cosine transform device is provided to commonly use multipliers, adders and subtracters composing two 1D IDCT for realizing a 2D IDCT, thereby reducing the entire area by half and errors. CONSTITUTION: An inverse discrete cosine transform device includes first and second input terminals(41a,41b) receiving two pixel data having different bits, a pre-processor(42) for equalizing the bits by compensating the shorter bit width by a bit difference, first and second MUX parts(43,46), a calculation logic operation part(44), first and second post processors(45), and a shifter(47). The first MUX part either one of the input pixel data according to first or second control value input from the outside. The calculation logic operation part has a multiplier/adder(44a) and an adder/subtracter(44b) for operating an output value from the first MUX part. The first and second post processing parts compensates the output value from the calculation logic operation part. The second MUX part outputs first and second 1D IDCT according to the first or second control value from the values output from the first and second post processors. The shifter shifts the output value from the second MUX part by a number2¬(n-m) smaller than a number(n), wherein n is the number multiplied to a coefficient of the multiplier of the operation part, and m is smaller than n, both numbers n & m are integers. In the case of the second 1D IDCT, the shifting is carried out by 2¬(n+m).
摘要翻译：目的：提供一种逆离散余弦变换装置，用于共同使用构成两个1D IDCT的乘法器，加法器和减法器来实现2D IDCT，从而将整个区域减少一半和错误。构成：逆离散余弦变换装置包括接收具有不同位的两个像素数据的第一和第二输入端（41a，41b），用于通过以比特差补偿较短位宽度来均衡位的预处理器（42），第一和第二MUX部分（43,46），计算逻辑运算部分（44），第一和第二后处理器（45）和移位器（47）。第一MUX部分根据从外部输入的第一或第二控制值的输入像素数据中的任一个。计算逻辑运算部分具有用于操作来自第一MUX部分的输出值的乘法器/加法器（44a）和加法器/减法器（44b）。第一和第二后处理部件从计算逻辑运算部分补偿输出值。第二MUX部分根据从第一和第二后处理器输出的值根据第一或第二控制值输出第一和第二1D IDCT。移位器将来自第二MUX部分的输出值移位小于数字（n）的数字2（nm），其中n是与操作部分的乘数的系数相乘的数字，并且m小于n，两个数字n＆m都是整数。在第二1DIDCT的情况下，移位由2π（n + m）进行。

52. 发明公开

KR1020030019787A 행렬 변화를 통한 분산산술처리 방식의 이산여현변환 방법 有权
标题翻译：通过矩阵转换的分布式算法离散COSE变换方法
公开(公告)号：KR1020030019787A
公开(公告)日：2003-03-07
申请号：KR1020010053106
申请日：2001-08-31
申请人： 주식회사 아이에이
发明人： 양현수
IPC分类号： H04N19/42 , H04N19/625
CPC分类号： H04N19/436 , H04N19/625
摘要： PURPOSE: A distributed arithmetic discrete cosine transform method through matrix conversion is provided to change a matrix of discrete cosine transform to reduce the size of hardware, decrease the quantity of calculations and increase accuracy in calculations. CONSTITUTION: Forward or backward discrete cosine transform is performed on the basis of a matrix calculation expression that is obtained by multiplying both sides of a one-dimensional DCT or IDCT matrix calculation expression by a predetermined number. The discrete cosine transform shifts one bit to the right after the first one-dimensional calculation, and shifts two bits to the right after the second one-dimensional calculation, to compensate the number multiplied by both sides of the matrix.
摘要翻译：目的：提供一种通过矩阵转换的分布式算术离散余弦变换方法，以改变离散余弦变换矩阵，减少硬件尺寸，减少计算量，提高计算精度。构成：基于通过将一维DCT或IDCT矩阵计算表达式的两边乘以预定数量而获得的矩阵计算表达式执行前向或后向离散余弦变换。离散余弦变换在第一次一维计算之后向右移位一位，并且在第二次一维计算之后向右移位两位，以补偿乘以矩阵两侧的数字。

53. 发明授权

KR100329367B1 영상 압축 시스템에서 디씨티 구조 失效
标题翻译：图像压缩系统中的DCT结构
公开(公告)号：KR100329367B1
公开(公告)日：2002-03-22
申请号：KR1019990057535
申请日：1999-12-14
申请人： 엘지정보통신주식회사
发明人： 윤태인
IPC分类号： H04N19/42 , H04N19/625
CPC分类号： G06T9/007
摘要： 본발명은영상압축시스템(System)에서 DCT(Discrete Cosine Transform) 구조에관한것으로, 특히 DCT 알고리즘(Algorithm)에서곱셈연산을처리하는방식을고속직병렬 MAC(Media Access Control)로구현하도록한 영상압축시스템에서 DCT 구조에관한것이다. 본발명은입력데이터를비트순차형태로출력하는전처리프로세서를구비하는영상압축시스템에서디씨티구조에있어서, 2 개의입력데이터에대하여미리계산된정보를가지는다수개의 2 입력 LUT와 2 비트직렬 2 단파이프라인의 SCSA를구비하여상기전처리프로세서로부터비트순차형태의데이터를인가받아내적연산을수행한후에짝수번째데이터를출력하는다수개의짝수 DAU와; 4 개의입력데이터에대하여미리계산된정보를가지는다수개의 4 입력 LUT와 2 비트직렬 2 단파이프라인의 SCSA를구비하여상기전처리프로세서로부터비트순차형태의데이터를인가받아내적연산을수행한후에홀수번째데이터를출력하는홀수 DAU를포함하여이루어진것을특징으로한다.

54. 发明公开

KR1020010056102A 영상 압축 시스템에서 디씨티 구조 失效
标题翻译：视频压缩系统中的DCT结构
公开(公告)号：KR1020010056102A
公开(公告)日：2001-07-04
申请号：KR1019990057535
申请日：1999-12-14
申请人： 엘지정보통신주식회사
发明人： 윤태인
IPC分类号： H04N19/42 , H04N19/625
CPC分类号： G06T9/007
摘要： PURPOSE: A DCT structure in a video compression system is provided to increase data processing speed by designing an efficient 2-bit 2-stage pipeline SCSA, by reducing a carry propagation delay time, thereby processing data in the form of perfect 2-bit order. CONSTITUTION: A DCT(Discrete Cosine Transform) structure is composed of a pre-processor(20) and a DAU(Distributed Arithmetic Unit,30). The pre-processor(20) includes 12 adders. In addition, the pre-processor(20) receives data through an input port, and outputs the data in the form of bit order to the DAU(30). The DAU(30) performs an internal calculation between a fixed cosine coefficient and data outputted from the pre-processor(20) in the form of bit order. 4 even DAUs(40) receives the data processed from the pre-processor(20) and outputs even numbered data. 4 odd DAUs(50) receives the data processed from the pre-processor(20), and outputs odd numbered data.
摘要翻译：目的：提供视频压缩系统中的DCT结构，通过设计一个高效的2位2级流水线SCSA，通过减少进位传播延迟时间来提高数据处理速度，从而以完美的2位顺序处理数据。构成：DCT（离散余弦变换）结构由预处理器（20）和DAU（分布式算术单元，30）组成。预处理器（20）包括12个加法器。此外，预处理器（20）通过输入端口接收数据，并以位顺序的形式将数据输出到DAU（30）。 DAU（30）以固定余弦系数和以预处理器（20）输出的数据之间以位顺序的形式进行内部计算。 4个DAU（40）接收从预处理器（20）处理的数据并输出偶数数据。 4个DAU（50）接收从预处理器（20）处理的数据，并输出奇数数据。

55. 发明授权

KR100281685B1 압축 영상의 이동객체 분할장치 및 방법 失效
标题翻译：运动物体分割装置和压缩图像的方法
公开(公告)号：KR100281685B1
公开(公告)日：2001-02-15
申请号：KR1019980048436
申请日：1998-11-12
申请人： 한국전자통신연구원
发明人： 김재곤 , 이규원 , 김진웅
IPC分类号： G06T7/20 , H04N19/51 , H04N19/625
摘要： 본 발명은 영상 압축표준인 MPEG-1,2(Moving Picture Experts Group-1,2)에 의해 압축된 감시영상을 객체 기반으로 색인화하기 위한 이동객체 분할장치 및 방법에 관한 것이다.
이러한 본 발명은, 프레임 단위의 압축 영상의 비트열을 부분 복호하여 각 블록마다 이산여현변환(Discrete Cosine Transform)의 DC계수를 추출하여 가로, 세로 각각 8배로 축소된 DC영상으로 재구성하는 DC영상 재구성부와, 상기 재구성된 DC영상을 이용하여 움직임 객체를 분할하는 DC영상 분할부, 압축 영상의 비트열을 부분 복호하여 움직임 벡터를 추출하는 움직임 벡터 추출부, 및 상기 움직임 벡터를 이용하여 움직임 객체 중에서 잘못 추출된 움직임 객체를 제거하는 시-공간 분할부로 구성된 이동객체 분할장치를 제공한다.

56. 发明公开

KR1020010007522A 이산 코사인 변환 계수 영역에 정보를 삽입하는 시스템 및방법 失效
标题翻译：将信息插入DCT系数的系统和方法
公开(公告)号：KR1020010007522A
公开(公告)日：2001-01-26
申请号：KR1020000035047
申请日：2000-06-24
申请人： 닛뽕덴끼 가부시끼가이샤
发明人： 하시모또마사히로
IPC分类号： H04N19/625 , H04N19/46
CPC分类号： H04N19/18 , H04N19/115 , H04N19/124 , H04N19/129 , H04N19/149 , H04N19/15 , H04N19/176 , H04N19/46 , H04N19/60 , H04N19/91 , H04N19/625
摘要： PURPOSE: To insert a large amount of information into a DCT coefficient being an intermediate data form in a digital picture encoding system following an MPEG standard by providing a total code length restoration means making the total code length of a code generated by the DCT coefficient to be equal to total code length before information is inserted. CONSTITUTION: An information insertion means 1 inserts additional information into original information in one block by changing the value of one or not less than two DCT coefficient from among all DCT coefficients in one block. Namely, the information insertion means 1 inserts information, which is supplied from an insertion information storage means 3 and which is to be inserted into the DCT coefficient of inputted picture data. A total code length restoration means 5 refers to a variable length code table 7 and rewrites a part of DCT coefficient. Thus, total code length, when the DCT coefficient after information is inserted is encoded, is adjusted so as to be equal to total code length when the DCT coefficient prior to information is inserted, is encoded.
摘要翻译：目的：通过提供总代码长度恢复装置，将由DCT系数产生的代码的总代码长度设置为总数码长度恢复装置，将大量的信息插入到作为MPEG标准的数字图像编码系统中的中间数据形式的DCT系数中等于插入信息之前的总代码长度。构成：信息插入装置1通过从一个块中的所有DCT系数中改变一个或不少于两个DCT系数的值，将附加信息插入到一个块中的原始信息中。也就是说，信息插入装置1插入从插入信息存储装置3提供的要被插入的输入图像数据的DCT系数的信息。总码长恢复装置5参考可变长度码表7并重写DCT系数的一部分。因此，编码在信息插入后的DCT系数被编码时的总代码长度被编码为当信息插入之前的DCT系数被编码时等于总代码长度。

57. 发明公开

KR1020000050286A 이산 코사인 변환에 의해 부호화 된 영상에 생긴 블록현상의 제거 방법 无效
标题翻译：通过使用离散COSINE变换解除图像的阻塞效应的方法
公开(公告)号：KR1020000050286A
公开(公告)日：2000-08-05
申请号：KR1019990000019
申请日：1999-01-02
申请人： 삼성전자주식회사
发明人： 지종완
IPC分类号： H04N19/625 , H04N19/86
CPC分类号： H04N19/86 , H04N19/176 , H04N19/625
摘要： PURPOSE: A method for removing blocking effect on an image decoded by using discrete cosine transformation is provided to remove blocking effects. CONSTITUTION: A method for removing blocking effects on an image decoded by using discrete cosine transformation includes following steps. At the first step, a macro block(20) of 2NX2N size(N=8) is formed for each decoded image. At the second step, the macro blocks are arranged so as not to overlap one another. At the third step, a first low band filter process is performed by performing discrete cosine transformation on each macro block. At the forth step, a super macro block(30) of 4NX4N type(N=8) ks formed. At the fifth step, the super macro blocks are arranged so as not to overlap one another. At the sixth step, a second low band filter process is performed by performing discrete cosine transformation on each super macro block.
摘要翻译：目的：提供一种通过使用离散余弦变换解码图像去除阻塞效应的方法，以消除阻塞效应。构成：通过使用离散余弦变换解码的图像的去除遮挡效应的方法包括以下步骤。在第一步骤中，为每个解码图像形成2NX2N大小（N = 8）的宏块（20）。在第二步骤中，宏块被布置为不彼此重叠。在第三步骤中，通过对每个宏块执行离散余弦变换来执行第一低频带滤波处理。在第四步，形成4NX4N型（N = 8）ks的超宏块（30）。在第五步骤中，超宏块被布置为不彼此重叠。在第六步骤中，通过对每个超级宏块执行离散余弦变换来执行第二低频带滤波处理。

58. 发明公开

KR1020000021859A 영상부호화장치 失效
标题翻译：编码图像的装置
公开(公告)号：KR1020000021859A
公开(公告)日：2000-04-25
申请号：KR1019980041137
申请日：1998-09-30
申请人： 대우전자주식회사
发明人： 한석원
IPC分类号： H04N19/17 , H04N19/625
CPC分类号： H04N19/176 , H04N19/625
摘要： PURPOSE: An apparatus for encoding an image is provided to estimate a DCT type and revival CBPY information without an error using the estimated DCT type. CONSTITUTION: An apparatus for encoding an image includes a shape encoder(110), a shape inspection part(100), DCT type decis ion part(120), and texture encoder(180). The shape encoder is to generate an encoded form information, decode an encoded form information and generate a revival form information. The shape inspection part is to provide a block designation signal deciding a block type of processing macro block and presenting a block type using the revival form information. The block type is composed of an object block comprising an object pixel only, a background block comprising background pixel only, and a boundary block comprising the background pixel and, at least, one or more object pixel. The DCT type decis ion part is to decide a DCT type presenting whether encodes the texture information using a sequential encoding method or a reverse encoding method or not depending on the block designation signal and the texture information and provide the decided DCT type information. The texture encoder is to encode the texture information using the sequential encoding method or the reverse encoding method responding the DCT type information.
摘要翻译：目的：提供一种用于编码图像的装置，以使用估计的DCT类型来估计DCT类型和不产生误差的复原CBPY信息。构成：用于编码图像的装置包括形状编码器（110），形状检查部件（100），DCT型判定部件（120）和纹理编码器（180）。形状编码器是生成编码形式信息，对编码形式信息进行解码并生成复制形式信息。形状检查部件提供决定块类型的处理宏块的块指定信号，并使用复制形式信息呈现块类型。块类型由仅包括对象像素的对象块，仅包含背景像素的背景块和包括背景像素的边界块以及至少一个或多个对象像素组成。 DCT类型决定部分是根据块指定信号和纹理信息来确定使用顺序编码方法或反向编码方法来编码纹理信息的DCT类型，并提供所确定的DCT类型信息。纹理编码器使用响应DCT类型信息的顺序编码方法或反向编码方法对纹理信息进行编码。

59. 发明公开

KR1020000018323A 화상압축장치와방법및화상복원장치. 失效
标题翻译：图像压缩装置和图像重构装置
公开(公告)号：KR1020000018323A
公开(公告)日：2000-04-06
申请号：KR1019980035862
申请日：1998-09-01
申请人： 주식회사 엘지이아이
发明人： 김정민 , 이상용
IPC分类号： H04N19/66 , H04N19/625
CPC分类号： H04N19/66 , H04N19/48 , H04N19/625
摘要： PURPOSE: An image compress apparatus is provided to maintain a quality of a source image data to be transferred. CONSTITUTION: The image compress apparatus comprises: a discrete cosine transform(DCT) executing part (21) for frequency-converting an image divided into a plurality of blocks; a quantization executing part (22) for quantizing the frequency-converted data using a constant reference quantization value; a scanning part (23) for rearranging the quantization data of each block of in a transfer unit into the same frequency, respectively; and a coding part (24) for coding the scanned data.
摘要翻译：目的：提供图像压缩装置以保持要传送的源图像数据的质量。构成：图像压缩装置包括：离散余弦变换（DCT）执行部分（21），用于对分成多个块的图像进行频率转换; 用于使用恒定的参考量化值量化所述频率转换数据的量化执行部分（22）; 扫描部分（23），用于将传送单元中的每个块的量化数据分别重新排列成相同的频率; 以及用于对扫描数据进行编码的编码部分（24）。

60. 发明授权

KR100246878B1 역이산 코사인변환 프로세서, 이산 코사인 변환계수의 매트릭스를 역변환하기 위한 방법 및 그 장치 失效
标题翻译：反向离散COSINE变换处理器
公开(公告)号：KR100246878B1
公开(公告)日：2000-03-15
申请号：KR1019940008377
申请日：1994-04-19
申请人： 지아이 코포레이션
发明人： 크리스토퍼엘.후겐붐
IPC分类号： H04N19/625
CPC分类号： G06F17/147
摘要： 본 발명에 따르면, 이산 코사인변환(DCT) 계수를 역변환하기 위한 방법 및 장치가 제공된다. DCT계수는 DCT케넬 매트릭스에 따라 벡터 도트 프로덕트를 얻도록 한번에 적어도 2비트를 처리한다. 벡터 도트 프로덕트의 매트릭스는 도트 프로덕트 엘리먼트의 전치 매트릭스를 얻도록 전치된다. 전치 매트릭스의 도트 프로덕트 엘리먼트는 DCT케넬 매트릭스에 따라 벡터 도트 프로덕트를 얻도록 한번에 적어도 2비트를 처리한다. 결과적인 도트 프로덕트가 적절한 순서로 인터리브됨과 더불어 일련의 출력 레지스터에 버퍼된다. 다른 효과가 출력 레지스터 이전에 ICDT흐름그래프의 후가산단을 구현함으로써 달성된다.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式