公开(公告)号：US5633685A

公开(公告)日：1997-05-27

申请号：US629173

申请日：1996-04-08

申请人： Soon-Jae Cho ,  Jae-Gark Choi ,  Si-Woong Lee ,  Yi-Han Kim ,  Seong-Dae Kim ,  Hae-Mook Jung

发明人： Soon-Jae Cho ,  Jae-Gark Choi ,  Si-Woong Lee ,  Yi-Han Kim ,  Seong-Dae Kim ,  Hae-Mook Jung

IPC分类号： H04N11/04 ,  G06T9/00 ,  H03M7/30 ,  H04N7/24 ,  H04N19/102 ,  H04N19/132 ,  H04N19/136 ,  H04N19/14 ,  H04N19/176 ,  H04N19/18 ,  H04N19/189 ,  H04N19/196 ,  H04N19/20 ,  H04N19/423 ,  H04N19/50 ,  H04N19/503 ,  H04N19/59 ,  H04N19/61 ,  H04N19/625 ,  H04N7/30 ,  G06K9/00

CPC分类号： H04N19/649 ,  H04N19/20

摘要： A processing block is processed by using an extension-interpolation ("E-I") method, wherein the processing block consisting of N.times.N pixels which are classified into object pixels and background pixels and the E-I method comprises the steps of: (A) selecting L rows of the processing block, each having at least one object pixels; (B) for each selected row, determining M representing a number of object pixels included in said each selected row and providing a first vector representing the M object pixels; (C) for each first vector, selecting an extension matrix among a multiplicity of predetermined extension matrices based on the M and N values and multiplying the selected extension matrix to the first vector to provide an extended first vector; (D) providing a first extended processing block which includes L rows, each row representing each of the extended first vectors; (E) for each column of the first extended processing block, providing a corresponding L-dimensional second vector; (F) selecting an extension matrix among a multiplicity of the predetermined extension matrices based on the L and N values; and (G) multiplying the extension matrix selected at step (F) to each of the second vectors, to provide N extended second vectors.

摘要翻译： 通过使用扩展插值（“EI”）方法处理处理块，其中由分类为对象像素和背景像素的N×N个像素组成的处理块和EI方法包括以下步骤：（A）选择L行 处理块，每个具有至少一个对象像素; （B），确定M表示包括在所述每个所选行中的对象像素的数量，并提供表示所述M个对象像素的第一矢量; （C）对于每个第一向量，基于M和N值在多个预定扩展矩阵中选择扩展矩阵，并将所选择的扩展矩阵与第一向量相乘以提供扩展的第一向量; （D）提供包括L行的第一扩展处理块，每行表示每个扩展的第一向量; （E），提供相应的L维第二向量; （F）基于L和N值在多个预定扩展矩阵中选择扩展矩阵; 和（G）将在步骤（F）选择的扩展矩阵与每个第二向量相乘，以提供N个扩展的第二向量。

公开(公告)号：US5734737A

公开(公告)日：1998-03-31

申请号：US502228

申请日：1995-07-13

申请人： Gyu-Hwan Chang ,  Hae-Mook Jung ,  Seong-Dae Kim ,  Jae-Gark Choi ,  Si-Woong Lee ,  Soon-Jae Cho

发明人： Gyu-Hwan Chang ,  Hae-Mook Jung ,  Seong-Dae Kim ,  Jae-Gark Choi ,  Si-Woong Lee ,  Soon-Jae Cho

IPC分类号： G06T7/20 ,  H04N19/00 ,  H04N19/105 ,  H04N19/139 ,  H04N19/147 ,  H04N19/182 ,  H04N19/192 ,  H04N19/196 ,  H04N19/20 ,  H04N19/503 ,  H04N19/51 ,  G06T9/00

CPC分类号： G06T7/2053 ,  H04N19/53

摘要： A method, for use in an object-oriented analysis-synthesis coder, for segmenting an image signal into moving objects contained in a current frame and describing each motion of the moving objects with a set of motion parameters using a multi-stage segmentation comprises the steps of: (a) comparing the current frame with its preceding frame to detect a moving area in the current frame; (b) further comparing the current frame with the preceding frame to detect a motion vector for each pixel in the moving area, i.e., a motion vector field; (c) first segmenting the motion vector field by using a two dimensional translational patch model; (d) secondly segmenting the first stage segmenting results by using a planar patch model; and (e) thirdly segmenting the second-stage segmenting results by using a parabolic patch model.

摘要翻译： 一种用于面向对象的分析合成编码器中的方法，用于将图像信号分割成包含在当前帧中的移动对象，并且使用多级分段来描述运动对象的一组运动参数的每个运动，包括： 步骤：（a）将当前帧与其前一帧进行比较，以检测当前帧中的移动区域; （b）进一步比较当前帧与前一帧以检测运动区域中的每个像素即运动矢量场的运动矢量; （c）通过使用二维平移补片模型首先分割运动矢量场; （d）使用平面贴片模型二次分割第一阶段分割结果; 和（e）通过使用抛物线贴片模型第三次分割第二阶段分割结果。

公开(公告)号：US5748761A

公开(公告)日：1998-05-05

申请号：US502229

申请日：1995-07-13

申请人： Gyu-Hwan Chang ,  Hae-Mook Jung ,  Seong-Dae Kim ,  Jae-Gark Choi ,  Si-Woong Lee

发明人： Gyu-Hwan Chang ,  Hae-Mook Jung ,  Seong-Dae Kim ,  Jae-Gark Choi ,  Si-Woong Lee

IPC分类号： H04N19/105 ,  H04N7/24 ,  H04N19/134 ,  H04N19/136 ,  H04N19/137 ,  H04N19/139 ,  H04N19/167 ,  H04N19/176 ,  H04N19/189 ,  H04N19/192 ,  H04N19/196 ,  H04N19/20 ,  H04N19/50 ,  H04N19/503 ,  H04N19/51 ,  H04N19/513 ,  H04N19/533 ,  H04N19/537 ,  H04N19/56 ,  H04N19/57 ,  G06T7/20

CPC分类号： H04N19/543 ,  H04N19/20

摘要： A method for use in an object-oriented analysis-synthesis coder, for segmenting an image of a current frame into moving objects and describing each motion of the moving objects with a set of motion parameters, which comprises the steps of: (a) comparing the current frame with a preceding frame to detect a motion vector field; (b) sliding window blocks on the motion vector field, and computing the degree of homogeneity for each of the window blocks to produce a seed block; (c) determining a set of initial values for the set of motion parameters based on the seed block; (d) evaluating whether each motion vector contained in and around the seed block is describable with the set of initial values or not, to thereby detect a revised region; (e) determining a set of revised values for the set of motion parameters from the revised region; (f) evaluating whether each motion vector contained in and around the revised region is describable with the set of revised values or not, to thereby detect a newly revised region; and (g) repeating the steps (e) and (f) until the difference between a current revised region and a previous revised region is less than a predetermined value, to thereby determine a set of final values for the set of motion parameters.

摘要翻译： 一种在面向对象分析合成编码器中使用的方法，用于将当前帧的图像分割成运动对象，并用一组运动参数描述运动对象的每个运动，其包括以下步骤：（a）比较 当前帧具有前一帧以检测运动矢量场; （b）在运动矢量场上滑动窗口块，并且计算每个窗口块的均匀度以产生种子块; （c）基于所述种子块确定所述一组运动参数的一组初始值; （d）评估包含在种子块内和周围的每个运动矢量是否可以用初始值的集合来描述，从而检测修正的区域; （e）从经修订的区域确定一组运动参数的修正值; （f）评估修改后区域内及其周围的每个运动矢量是否可以用一组修订值进行描述，从而检测新修订的区域; 和（g）重复步骤（e）和（f），直到当前修订区域和先前修订区域之间的差值小于预定值，从而确定该组运动参数的一组最终值。

公开(公告)号：US5682341A

公开(公告)日：1997-10-28

申请号：US435912

申请日：1995-04-19

申请人： Seong-Dae Kim ,  Tae-Sung Kim

发明人： Seong-Dae Kim ,  Tae-Sung Kim

IPC分类号： G06F17/11 ,  G06F17/10

CPC分类号： G06F17/11

摘要： An adaptive signal processor using a Newton/least mean square (LMS) algorithm, adapted to remove a signal distortion or noise. The adaptive signal processor is adapted to derive an autocorrelation matrix from an input signal, derive an inverse matrix of the autocorrelation matrix by executing DCT twice or executing DFT and IDFT, and apply the derived inverse matrix to an adaptive signal processing, thereby capable of greatly reducing the total amount of computations to derive the inverse matrix, achieving an easy hardware realization thereof and reducing the manufacture cost thereof.

摘要翻译： 一种使用牛顿/最小均方（LMS）算法的自适应信号处理器，适用于消除信号失真或噪声。 自适应信号处理器适于从输入信号导出自相关矩阵，通过执行DCT或执行DFT和IDFT来导出自相关矩阵的逆矩阵，并将导出的逆矩阵应用于自适应信号处理，从而能够大大地 减少计算的总量以导出逆矩阵，实现其轻松的硬件实现并降低其制造成本。

公开(公告)号：US5502510A

公开(公告)日：1996-03-26

申请号：US308238

申请日：1994-09-19

申请人： Jong-Hoon Kim ,  Seong-Dae Kim

发明人： Jong-Hoon Kim ,  Seong-Dae Kim

IPC分类号： H04N19/50 ,  G06T9/00 ,  H04N5/21 ,  H04N19/102 ,  H04N19/117 ,  H04N19/139 ,  H04N19/189 ,  H04N19/423 ,  H04N19/426 ,  H04N19/51 ,  H04N19/625 ,  H04N19/80 ,  H04N19/85 ,  H04N7/24

CPC分类号： H04N19/80 ,  H04N5/21

摘要： A motion adaptive spatial filtering method is employed as a prefilter in an image coding apparatus, which processes the temporal band-limitation of a video signal without temporal aliasing by using a spatial filter having a band-limitation characteristic according to a desired temporal cutoff frequency.

摘要翻译： 在图像编码装置中采用运动自适应空间滤波方法作为预滤波器，其通过使用具有根据期望的时间截止频率的带限特性的空间滤波器来处理视频信号的时间频带限制而没有时间混叠。

公开(公告)号：US5502489A

公开(公告)日：1996-03-26

申请号：US176794

申请日：1994-01-03

申请人： Jong-Hun Kim ,  Seong-Dae Kim

发明人： Jong-Hun Kim ,  Seong-Dae Kim

IPC分类号： H04N7/12 ,  H04N7/26 ,  H04N7/18

CPC分类号： H04N19/80 ,  H04N7/122

摘要： The present invention is a motion adaptive spatial filtering(MASF) method for use with a image coding apparatus, which proceeds the temporal band-limitation of the video frame signals on spatial domain along the trajectory of a moving component without temporal aliasing by using a filter having a band-limitation characteristic according to a cutoff frequency. It is well known that video signal may contain a lot of temporal aliasing components because of insufficient frame sampling rate. And it is impossible to design band-limiting temporal filter simply by convolving a sequence of frames along the temporal direction. From a signal processing point of view, the aliasing degrades filter characteristics. This filtering method takes advantage of the fact that, if motion of each pixel is known, the temporal filtering can be performed in the spatial domain filtering along its trajectory. That results in de-aliasing characteristic.

摘要翻译： 本发明是一种与图像编码装置一起使用的运动自适应空间滤波（MASF）方法，其通过使用滤波器沿着运动分量的轨迹在空间域上进行视频帧信号的时间带限，而没有时间混叠 具有根据截止频率的频带限制特性。 众所周知，由于帧采样率不足，视频信号可能包含大量的时间混叠分量。 简单地通过沿着时间方向卷积帧序列来设计带限制时间滤波是不可能的。 从信号处理的角度来看，混叠降低了滤波器特性。 这种滤波方法利用了如下事实：如果每个像素的运动是已知的，则可以沿着其轨迹在空间域滤波中执行时间滤波。 这导致去锯齿特性。