公开(公告)号：US4903128A

公开(公告)日：1990-02-20

申请号：US221453

申请日：1988-08-25

申请人： Dominique Thoreau

发明人： Dominique Thoreau

IPC分类号： G06T9/00 ,  H03M7/30 ,  H04N1/41 ,  H04N7/26 ,  H04N7/30

CPC分类号： H04N19/98 ,  H04N19/86

摘要： Disclosed is a method for the reduction of the noise generated during the encoding/decoding of digitized images operating by blocks. To reduce the faults visible in the restored image, the method consists in filtering the image by using a filtering window, for example with dimensions of 3.times.3 pixels, which scans the image, the current point being the central point of this window, and in determining, for each point, the mean of the values associated wih the points of this window, and in determining for the current point, from this mean and from a characteristic value of the operating status of the encoder, a filtering parameter which can be used to determine the filtered value associated with the current point. The invention can be applied to the resutitution of block encoded images especially in digitalized video tape-recorders.

摘要翻译： PCT No.PCT / FR87 / 00425 Sec。 371日期1988年8月25日第 102（e）日期1988年8月25日PCT提交1987年10月28日PCT公布。 出版物WO88 / 03346 日期：1988年5月5日。公开的是减少在按块操作的数字化图像的编码/解码期间产生的噪声的方法。 为了减少恢复的图像中可见的故障，该方法包括通过使用过滤窗口过滤图像，例如尺寸为3x3像素，扫描图像，当前点是该窗口的中心点，并确定 对于每个点，与该窗口的点相关联的值的平均值，并且根据该平均值和从编码器的操作状态的特征值确定当前点，可以将其用于 确定与当前点相关联的过滤值。 本发明可以应用于块编码图像的替换，特别是在数字化录像机中。

公开(公告)号：US09554056B2

公开(公告)日：2017-01-24

申请号：US13618403

申请日：2012-09-14

申请人： David Touze ,  Yannick Olivier ,  Philippe Bordes ,  Dominique Thoreau ,  Joan Llach

发明人： David Touze ,  Yannick Olivier ,  Philippe Bordes ,  Dominique Thoreau ,  Joan Llach

IPC分类号： H04N7/12 ,  H04N5/235 ,  H04N5/355 ,  H04N19/70 ,  H04N19/46 ,  H04N19/33 ,  H04N19/98 ,  H04N19/59 ,  H04N5/345

CPC分类号： H04N5/2355 ,  H04N5/2356 ,  H04N5/3456 ,  H04N5/355 ,  H04N19/33 ,  H04N19/46 ,  H04N19/59 ,  H04N19/70 ,  H04N19/98

摘要： The invention is made in the field of coding of images of high dynamic range.The invention is based on the concept of Frame Compatible format. The idea is to transport, in a frame, down-sampled LDR content together with additional information allowing reconstructing HDR content from the LDR content.Thus, it is proposed a method of encoding an HDR image of high dynamic range according to claim 1. Said method comprises down-sampling (DWN) an LDR image and additional data, the LDR image providing a lower dynamic range depiction of the HDR image content and the additional data allowing for reconstructing the HDR image from the LDR image.

摘要翻译： 因此，提出了根据权利要求1对高动态范围的HDR图像进行编码的方法。所述方法包括下采样（DWN）LDR图像和附加数据，LDR图像提供HDR图像的较低动态范围描绘 内容和允许从LDR图像重建HDR图像的附加数据。

公开(公告)号：US08428132B2

公开(公告)日：2013-04-23

申请号：US12308907

申请日：2007-06-26

申请人： Philippe Guillotel ,  Dominique Thoreau ,  Julien Haddad

发明人： Philippe Guillotel ,  Dominique Thoreau ,  Julien Haddad

IPC分类号： H04N7/12

CPC分类号： H04N19/85 ,  H04N19/112 ,  H04N19/137 ,  H04N19/14 ,  H04N19/159 ,  H04N19/172 ,  H04N19/176 ,  H04N19/61 ,  H04N19/80

摘要： The method comprises the following steps: calculation of the spatial activity of each macroblock and of the temporal activity between the macroblocks constituting a group of macroblocks in field mode, calculation of an activity index relating to the group that is an increasing function of the temporal activity and a decreasing function of the spatial activity of the group, comparison of the activity index relating to the group, to a predefined threshold, to define it as a group having activity, coding of the picture in field or frame mode according to the number of macroblock groups detected having activity with respect to the number of groups of macroblocks in the picture. The application relate to data compression, particularly the H 264 standard.

公开(公告)号：US08416992B2

公开(公告)日：2013-04-09

申请号：US11795001

申请日：2006-01-10

申请人： Olivier Le Meur ,  Dominique Thoreau ,  Philippe Salmon ,  Dominique Barba ,  Patrick Le Callet

发明人： Olivier Le Meur ,  Dominique Thoreau ,  Philippe Salmon ,  Dominique Barba ,  Patrick Le Callet

IPC分类号： G06K9/00

CPC分类号： G06K9/4671 ,  G06K9/4623 ,  G06K9/4652 ,  G06K9/629 ,  G06T7/262

摘要： Detection of the salient points in an image enable the improvement of further steps such as coding or image indexing, watermarking, video quality estimation. The methods rely on the fact that a model is fully based on the human visual system (HVS) such as the computation of early visual features, and the methods compute a saliency map for video images taking into account motion and the velocity of the eye.

摘要翻译： 检测图像中的突出点使得能够改进进一步的步骤，例如编码或图像索引，水印，视频质量估计。 该方法依赖于一个模型完全基于人类视觉系统（HVS），例如早期视觉特征的计算，并且该方法计算视频图像的显着图，考虑运动和眼睛的速度。

公开(公告)号：US08285064B2

公开(公告)日：2012-10-09

申请号：US12452694

申请日：2008-07-15

申请人： Julien Haddad ,  Dominique Thoreau ,  Philippe Salmon ,  Xavier Castellan ,  Christophe Chevance

发明人： Julien Haddad ,  Dominique Thoreau ,  Philippe Salmon ,  Xavier Castellan ,  Christophe Chevance

IPC分类号： G06K9/36

CPC分类号： H04N19/895 ,  H04N19/44 ,  H04N19/48 ,  H04N19/61 ,  H04N19/80 ,  H04N19/86

摘要： The invention concerns a method for processing images, texture pattern blocks representative of a typical texture of the images being stored in a memory, the method comprising the following steps: transforming the pixel grey levels of a current block into a current block of frequency coefficients using a predefined transform; building up a merger block, each frequency coefficient positioned at a defined place of the merger block being equal either to the frequency coefficient positioned at the same place of the current block, or to the frequency coefficient positioned at the same place of a texture pattern block among the texture pattern blocks stored in the memory; selecting a merger block according to a predefined function (J); and replacing in the image the current block by the selected merger block. The invention also concerns an electronic device for processing the corresponding images.

摘要翻译： 本发明涉及一种用于处理图像的方法，代表存储在存储器中的图像的典型纹理的纹理图案块，该方法包括以下步骤：将当前块的像素灰度级变换成当前的频率系数块，使用 预定义的变换; 建立合并块，位于合并块的定义位置的每个频率系数等于位于当前块的相同位置处的频率系数，或者与位于纹理图案块的相同位置处的频率系数相等 存储在存储器中的纹理图形块之间; 根据预定义的功能（J）选择合并块; 并且通过所选择的合并块在图像中替换当前块。 本发明还涉及用于处理相应图像的电子设备。

公开(公告)号：US08139881B2

公开(公告)日：2012-03-20

申请号：US11887714

申请日：2006-03-16

申请人： Lila Huguenel ,  Dominique Thoreau ,  Olivier Le Meur

发明人： Lila Huguenel ,  Dominique Thoreau ,  Olivier Le Meur

IPC分类号： G06K9/00

CPC分类号： H04N19/192 ,  H04N19/124 ,  H04N19/154 ,  H04N19/17 ,  H04N19/179 ,  H04N19/60

摘要： The invention relates to a method of locally adjusting a quantization step for each region of an image with respect to a setpoint quantization step. The method comprises the steps of: assigning, to said region of interest, a first quantization step, lower than or equal to said setpoint quantization step; determining a second quantization step higher than or equal to said setpoint quantization step and lower than or equal to a third quantization step such that the number of bits required by the region of interest is lower than the number of bits saved on the potential masking region; and assigning, when such a second quantization step exists, said second quantization step to said potential masking region, otherwise assigning said third quantization step to said potential masking region.

摘要翻译： 本发明涉及一种相对于设定点量化步骤对图像的每个区域进行局部调整量化步长的方法。 该方法包括以下步骤：向所述感兴趣区域分配低于或等于所述设定点量化步长的第一量化步长; 确定高于或等于所述设定点量化步长并低于或等于第三量化步长的第二量化步骤，使得所述感兴趣区域所需的位数低于保存在所述电位屏蔽区域上的位数; 以及当存在所述第二量化步骤时，将所述第二量化步骤分配给所述电位屏蔽区域，否则将所述第三量化步长分配给所述电位屏蔽区域。

公开(公告)号：US20090169125A1

公开(公告)日：2009-07-02

申请号：US11887860

申请日：2006-03-16

申请人： Lila Huguenel ,  Dominique Thoreau ,  Philippe Guillotel

发明人： Lila Huguenel ,  Dominique Thoreau ,  Philippe Guillotel

IPC分类号： G06K9/00

CPC分类号： H04N19/192 ,  H04N19/124 ,  H04N19/152 ,  H04N19/154 ,  H04N19/17 ,  H04N19/179 ,  H04N19/60

摘要： The invention relates to a method of locally adjusting a quantization step for each region of an image in a sequence of n images. The method comprises the following steps: calculating, for each image of the sequence, a first difference of bits saved on the potential masking region when quantizing it with a first quantization step and a second difference of bits required by the region of interest when quantizing it with a second quantization step; and in each image of the sequence, assigning to the potential masking region the first quantization step and if the sum of the n first differences of bits is higher than the sum of the n second differences of bits, assigning said second quantization step to said region of interest and otherwise assigning a third quantization step such that the sum of said n first differences of bits is distributed between the regions of interest according to a reconstruction quality criterion.

摘要翻译： 本发明涉及一种对n个图像序列中的图像的每个区域进行局部调整量化步长的方法。 该方法包括以下步骤：对于该序列的每个图像，当量化它时，以第一量化步长和感兴趣区域所需的位的第二差值，计算保存在潜在屏蔽区域上的位的第一差值， 用第二量化步骤; 并且在所述序列的每个图像中，将所述第一量化步骤分配给所述潜在掩蔽区域，并且如果所述n个第一差分比特的和高于所述n个第二差分比特的和，则将所述第二量化步长分配给所述区域 并且否则分配第三量化步骤，使得根据重构质量标准，所述n个第一差分的和被分布在感兴趣区域之间。

公开(公告)号：US07408987B1

公开(公告)日：2008-08-05

申请号：US09786432

申请日：1999-09-06

申请人： Christophe Chevance ,  Pierre Ruellou ,  Dominique Thoreau

发明人： Christophe Chevance ,  Pierre Ruellou ,  Dominique Thoreau

IPC分类号： H04N7/18

CPC分类号： H04N19/51 ,  H04N19/513 ,  H04N19/527

摘要： The method includes segmentation of the video image into image blocks, movement estimation per image block in order to obtain a field of movement vectors. It is characterized in that it includes a stage of reassignment of a vector to a block by selecting one movement vector from among N predominant vectors belonging to the field of vectors. The applications relate to movement estimation, for example, by image-block matching.

摘要翻译： 该方法包括将视频图像分割成图像块，每个图像块的运动估计，以获得运动矢量场。 其特征在于，它包括通过从属于矢量场的N个主要矢量中选择一个运动矢量来将向量重新分配给块的阶段。 应用涉及运动估计，例如通过图像块匹配。

公开(公告)号：US06580758B1

公开(公告)日：2003-06-17

申请号：US08137189

申请日：1994-03-14

申请人： Christophe Chevance ,  Dominique Thoreau

发明人： Christophe Chevance ,  Dominique Thoreau

IPC分类号： H04N718

CPC分类号： H04N19/98 ,  H04N19/593 ,  H04N19/61 ,  H04N19/82

摘要： A method and apparatus for differential coding of blocks of images includes coding the difference between a current block of an image to be transmitted and one of four blocks of the reconstructed current image in the closest vicinity of the current block. The device includes a differential image coding structure having a quantifier coupled to a variable-length coding device, a predictor and a device for reconstruction of the transmitted image coupled to a first image memory. The device further includes a second image memory for storing image blocks previously reconstructed in the image preceding the current image.

摘要翻译： 用于图像块的差分编码的方法和装置包括对要发送的图像的当前块与当前块最接近的重构的当前图像的四个块中的一个之间的差进行编码。 该装置包括差分图像编码结构，其具有耦合到可变长度编码装置的量化器，预测器和用于重建耦合到第一图像存储器的发射图像的装置。 该装置还包括第二图像存储器，用于存储先前在当前图像之前的图像中重构的图像块。

公开(公告)号：US08325804B2

公开(公告)日：2012-12-04

申请号：US12312117

申请日：2007-10-26

申请人： Julien Haddad ,  Dominique Thoreau ,  Philippe Guillotel

发明人： Julien Haddad ,  Dominique Thoreau ,  Philippe Guillotel

IPC分类号： H04N7/12

CPC分类号： H04N19/19 ,  H04N19/11 ,  H04N19/14 ,  H04N19/147 ,  H04N19/176 ,  H04N19/593 ,  H04N19/61

摘要： The method, for the determination of the coding mode of a macroblock constituted of picture blocks, carries out a pre-selection of intra directional predictive coding modes of macroblock blocks according to the following steps: calculation of the activities gradient of a block in the prediction directions, pre-selection of intra directional coding modes of a block for which the directions correspond to the activity or activities gradient of minimum value.

摘要翻译： 用于确定由图像块构成的宏块的编码模式的方法根据以下步骤执行宏块块的帧内预测编码模式的预选择：计算预测中的块的活动梯度 方向，预先选择方向对应于最小值的活动或活动梯度的块的内部方向编码模式。