公开(公告)号：US09088685B2

公开(公告)日：2015-07-21

申请号：US13640253

申请日：2011-04-07

申请人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： H04N5/217 ,  H04N9/04 ,  G06T5/00 ,  H04N9/64

CPC分类号： H04N9/045 ,  G06T5/002 ,  G06T2207/10024 ,  H04N9/646

摘要： The present invention provides a method for efficiently removing noise of an image sensor. To this end, the present invention enables: an area to be divided into an edge area, a sensitivity difference generation area, and a pattern area relative to an input image of a Bayer pattern; and different noise removing methods to be applied according to each of the divided areas. By doing this, noise can be removed efficiently since the noise is adaptively removed in each area. Further, sensitivity differences between Gr and Gb are considered in such a manner that resolution can be improved.

摘要翻译： 本发明提供了一种有效去除图像传感器噪声的方法。 为此，本发明能够：相对于拜耳图案的输入图像将区域划分为边缘区域，灵敏度差异生成区域和图案区域; 以及根据每个分割区域应用的不同的噪声去除方法。 通过这样做，可以有效地去除噪声，因为在每个区域中自适应地去除噪声。 此外，考虑到Gr和Gb之间的灵敏度差异，使得可以提高分辨率。

公开(公告)号：US08428388B2

公开(公告)日：2013-04-23

申请号：US13016171

申请日：2011-01-28

申请人： Han-Sae Song ,  Hee-Chan Park ,  Min-Kyu Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Han-Sae Song ,  Hee-Chan Park ,  Min-Kyu Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： G06K9/40

CPC分类号： G06T5/002 ,  G06T5/20 ,  G06T2207/20012 ,  G06T2207/20192

摘要： An image generating apparatus and method for emphasizing an edge, based on image characteristics by extracting an edge region from an input original image, and an edge gain indicating a degree for emphasizing the edge region and an edge threshold indicating a degree of complexity between pixels are calculated for each of a plurality of image regions of the original image. The extracted edge region is combined with the calculated edge gain. Thresholding is performed by adjusting the calculated edge threshold for the edge region combined with the edge gain, and the thresholding-applied edge region is combined with the original image to output a final image. By adaptively adjusting an edge gain and an edge threshold for a portion where an edge needs to be emphasized and the other portions in an image having multiple image characteristics, both edge emphasis and noise cancellation can be achieved together.

摘要翻译： 一种图像产生装置和方法，用于通过从输入的原始图像中提取边缘区域，以及指示用于强调边缘区域的程度的边缘增益和指示像素之间的复杂度的边缘阈值，基于图像特征来强调边缘， 为原始图像的多个图像区域中的每一个计算。 提取的边缘区域与计算的边缘增益组合。 通过调整与边缘增益结合的边缘区域的计算边缘阈值来执行阈值，并且将阈值应用边缘区域与原始图像组合以输出最终图像。 通过对边缘需要被强调的部分进行自适应地调整边缘增益和边缘阈值，并且具有多个图像特征的图像中的其他部分可以一起实现边缘加强和噪声消除。

公开(公告)号：US20110188774A1

公开(公告)日：2011-08-04

申请号：US13016171

申请日：2011-01-28

申请人： Han-Sae SONG ,  Hee-Chan Park ,  Min-Kyu Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Han-Sae SONG ,  Hee-Chan Park ,  Min-Kyu Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： G06K9/40

CPC分类号： G06T5/002 ,  G06T5/20 ,  G06T2207/20012 ,  G06T2207/20192

摘要： An image generating apparatus and method for emphasizing an edge, based on image characteristics by extracting an edge region from an input original image, and an edge gain indicating a degree for emphasizing the edge region and an edge threshold indicating a degree of complexity between pixels are calculated for each of a plurality of image regions of the original image. The extracted edge region is combined with the calculated edge gain. Thresholding is performed by adjusting the calculated edge threshold for the edge region combined with the edge gain, and the thresholding-applied edge region is combined with the original image to output a final image. By adaptively adjusting an edge gain and an edge threshold for a portion where an edge needs to be emphasized and the other portions in an image having multiple image characteristics, both edge emphasis and noise cancellation can be achieved together.

摘要翻译： 一种图像产生装置和方法，用于通过从输入的原始图像中提取边缘区域，以及指示用于强调边缘区域的程度的边缘增益和指示像素之间的复杂度的边缘阈值，基于图像特征来强调边缘， 为原始图像的多个图像区域中的每一个计算。 提取的边缘区域与计算的边缘增益组合。 通过调整与边缘增益结合的边缘区域的计算边缘阈值来执行阈值，并且将阈值应用边缘区域与原始图像组合以输出最终图像。 通过对边缘需要被强调的部分进行自适应地调整边缘增益和边缘阈值，并且具有多个图像特征的图像中的其他部分可以一起实现边缘加强和噪声消除。

公开(公告)号：US09111365B2

公开(公告)日：2015-08-18

申请号：US12853853

申请日：2010-08-10

申请人： Hee-Chan Park ,  Min-Kyu Park ,  Han-Sae Song ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Hee-Chan Park ,  Min-Kyu Park ,  Han-Sae Song ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： H04N5/335 ,  G06T3/40

CPC分类号： G06T3/4015 ,  G06T3/403

摘要： An edge-adaptive interpolation and noise filtering method is provided including performing primary interpolation on a first color, based on a first preset number of directions, with respect to a first color window obtained from an input image window, estimating an edge direction within a primary interpolated first color window obtained by the primary interpolation, based on a second preset number of directions, the second preset number of directions being larger than the first preset number, with respect to the primary interpolated first color window, performing secondary interpolation on the first color, based on the estimated edge direction, with respect to the primary interpolated first color window, and performing interpolation on a second color, based on the estimated edge direction, with respect to a second color window obtained from the input image window.

摘要翻译： 提供了一种边缘自适应内插和噪声滤波方法，包括：基于第一预设数量的方向，相对于从输入图像窗口获得的第一颜色窗口，对第一颜色执行主插值，估计主要内容中的边缘方向 基于相对于主内插第一颜色窗口的第二预设数量的方向，第二预定数量的方向大于第一预设数量的第一插值获得的内插第一颜色窗口，对第一颜色进行二次插值 基于估计边缘方向，相对于主内插第一颜色窗口，并且基于估计的边缘方向，相对于从输入图像窗口获得的第二颜色窗口，对第二颜色进行插值。

公开(公告)号：US20130027590A1

公开(公告)日：2013-01-31

申请号：US13640253

申请日：2011-04-07

申请人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： H04N5/217

CPC分类号： H04N9/045 ,  G06T5/002 ,  G06T2207/10024 ,  H04N9/646

摘要： The present invention provides a method for efficiently removing noise of an image sensor. To this end, the present invention enables: an area to be divided into an edge area, a sensitivity difference generation area, and a pattern area relative to an input image of a Bayer pattern; and different noise removing methods to be applied according to each of the divided areas. By doing this, noise can be removed efficiently since the noise is adaptively removed in each area. Further, sensitivity differences between Gr and Gb are considered in such a manner that resolution can be improved.

摘要翻译： 本发明提供了一种有效去除图像传感器噪声的方法。 为此，本发明能够：相对于拜耳图案的输入图像将区域划分为边缘区域，灵敏度差异生成区域和图案区域; 以及根据每个分割区域应用的不同的噪声去除方法。 通过这样做，可以有效地去除噪声，因为在每个区域中自适应地去除噪声。 此外，考虑到Gr和Gb之间的灵敏度差异，使得可以提高分辨率。

公开(公告)号：US20110221937A1

公开(公告)日：2011-09-15

申请号：US13045110

申请日：2011-03-10

申请人： Min-Kyu PARK ,  Han -Sae Song ,  Hee-Chan Park ,  Young -Kwon Yoon ,  Yong-Gu Lee

发明人： Min-Kyu PARK ,  Han -Sae Song ,  Hee-Chan Park ,  Young -Kwon Yoon ,  Yong-Gu Lee

IPC分类号： H04N9/64

CPC分类号： H04N9/045 ,  G06T5/002 ,  G06T5/20 ,  G06T2207/10024 ,  G06T2207/20032 ,  H04N9/646

摘要： A device and method for reducing color noise of an image by using a distance weight depending on a distance from a central pixel for each pixel of an image and an edge weight depending on the difference in luminance and chrominance with the central pixel is provided, which can effectively reduce the color noise of the image by using the correlation between the luminance and the chrominance and edge characteristics.

摘要翻译： 提供了一种通过使用依赖于与图像的每个像素的中心像素的距离的距离权重和依赖于与中心像素的亮度和色度的差异的边缘权重来减少图像的颜色噪声的装置和方法，其中 可以通过使用亮度和色度和边缘特性之间的相关性来有效地降低图像的颜色噪声。

公开(公告)号：US08837852B2

公开(公告)日：2014-09-16

申请号：US13021098

申请日：2011-02-04

申请人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： G06K9/00 ,  H04N5/367

CPC分类号： H04N5/3675 ,  H04N2209/046

摘要： A method for removing defective pixels using a signal processing scheme with a Bayer pattern, by determining an output value of the inter-channel according to whether or not a difference between a center pixel and an average of neighboring pixels, which have the same color as that of the center pixel, is greater than or equal to a threshold value; and using values of neighboring pixels nearest to a center pixel as an input, obtaining an output value of the cross-channel by finding a median value among a median value of vertical and horizontal lines including the center pixel, a median value of diagonal lines including the center pixel, and a value of the center pixel. Then, a defective pixel is removed, through the use of the output value of the inter-channel, the output value of the cross-channel, and the value of the center pixel.

摘要翻译： 一种使用具有拜耳模式的信号处理方案来去除缺陷像素的方法，通过根据中心像素与相邻像素的平均值之间的差是否确定相邻像素的差异来确定信道间的输出值，该方法具有与 中心像素的值大于或等于阈值; 并且使用最接近中心像素的相邻像素的值作为输入，通过在包括所述中心像素的垂直和水平线的中值中找到中值，获得所述交叉通道的输出值，所述对角线的中值包括 中心像素和中心像素的值。 然后，通过使用信道间的输出值，交叉通道的输出值和中心像素的值来去除缺陷像素。

公开(公告)号：US20110194764A1

公开(公告)日：2011-08-11

申请号：US13021098

申请日：2011-02-04

申请人： Min-Kyu PARK ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

发明人： Min-Kyu PARK ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Yong-Gu Lee

IPC分类号： G06K9/00

CPC分类号： H04N5/3675 ,  H04N2209/046

摘要： A method for removing defective pixels using a signal processing scheme with a Bayer pattern, by determining an output value of the inter-channel according to whether or not a difference between a center pixel and an average of neighboring pixels, which have the same color as that of the center pixel, is greater than or equal to a threshold value; and using values of neighboring pixels nearest to a center pixel as an input, obtaining an output value of the cross-channel by finding a median value among a median value of vertical and horizontal lines including the center pixel, a median value of diagonal lines including the center pixel, and a value of the center pixel. Then, a defective pixel is removed, through the use of the output value of the inter-channel, the output value of the cross-channel, and the value of the center pixel.

摘要翻译： 一种使用具有拜耳模式的信号处理方案来去除缺陷像素的方法，通过根据中心像素与相邻像素的平均值之间的差是否确定相邻像素的差异来确定信道间的输出值，该方法具有与 中心像素的值大于或等于阈值; 并且使用最接近中心像素的相邻像素的值作为输入，通过在包括所述中心像素的垂直和水平线的中值中找到中值，获得所述交叉通道的输出值，所述对角线的中值包括 中心像素和中心像素的值。 然后，通过使用信道间的输出值，交叉通道的输出值和中心像素的值来去除缺陷像素。

公开(公告)号：US08711251B2

公开(公告)日：2014-04-29

申请号：US13045110

申请日：2011-03-10

申请人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Youg-Gu Lee

发明人： Min-Kyu Park ,  Han-Sae Song ,  Hee-Chan Park ,  Young-Kwon Yoon ,  Youg-Gu Lee

IPC分类号： H04N5/217

CPC分类号： H04N9/045 ,  G06T5/002 ,  G06T5/20 ,  G06T2207/10024 ,  G06T2207/20032 ,  H04N9/646

摘要： A device and method for reducing color noise of an image by using a distance weight depending on a distance from a central pixel for each pixel of an image and an edge weight depending on the difference in luminance and chrominance with the central pixel is provided, which can effectively reduce the color noise of the image by using the correlation between the luminance and the chrominance and edge characteristics.

摘要翻译： 提供了一种通过使用依赖于与图像的每个像素的中心像素的距离的距离权重和依赖于与中心像素的亮度和色度的差异的边缘权重来减少图像的颜色噪声的装置和方法，其中 可以通过使用亮度和色度和边缘特性之间的相关性来有效地降低图像的颜色噪声。

公开(公告)号：US08411965B2

公开(公告)日：2013-04-02

申请号：US12718511

申请日：2010-03-05

申请人： Hee-Chan Park ,  Sung-Dae Cho ,  Min-Kyu Park ,  Jin-Ho Kim ,  Ji-Hye Kim ,  Yun-Je Oh

发明人： Hee-Chan Park ,  Sung-Dae Cho ,  Min-Kyu Park ,  Jin-Ho Kim ,  Ji-Hye Kim ,  Yun-Je Oh

IPC分类号： G06K9/48

CPC分类号： G06T7/13 ,  G06T2207/20021

摘要： An image processing method and apparatus capable of easily detecting an edge of an object from an input image, in which the edge is detected using one step, without a pre-processing step, and a complicated trigonometric function is not used for gradient detection. The image processing method includes setting a window within an input image, analyzing the window to determine directions of edges of objects within the image included in the window, detecting edge information including the edge directions, and processing and outputting the window using the edge information.

摘要翻译： 一种图像处理方法和装置，其不需要用于梯度检测，能够容易地从输入图像中检测出物体的边缘，其中使用一个步骤检测边缘，而无需预处理步骤，并且复杂的三角函数不用于梯度检测。 图像处理方法包括设置输入图像内的窗口，分析窗口以确定包括在窗口中的图像内的对象的边缘的方向，检测包括边缘方向的边缘信息，以及使用边缘信息来处理和输出窗口。