公开(公告)号：US5920654A

公开(公告)日：1999-07-06

申请号：US574572

申请日：1995-12-14

申请人： Jun Someya ,  Kazuhiro Chiba ,  Kazuyoshi Watabu ,  Yoshie Yamamoto ,  Mamoru Inamura ,  Takeshi Nagayama ,  Noriko Bamba ,  Noboru Sunada

发明人： Jun Someya ,  Kazuhiro Chiba ,  Kazuyoshi Watabu ,  Yoshie Yamamoto ,  Mamoru Inamura ,  Takeshi Nagayama ,  Noriko Bamba ,  Noboru Sunada

IPC分类号： G06T3/40 ,  G06T5/00 ,  G06T5/20 ,  H04N1/409 ,  H04N5/208 ,  H04N5/21 ,  H04N5/367 ,  H04N9/04 ,  G06K9/36

CPC分类号： G06T3/4007 ,  G06T5/004 ,  G06T5/20 ,  H04N1/409 ,  H04N5/208 ,  H04N5/21 ,  H04N5/367 ,  H04N9/045 ,  H04N2209/041 ,  H04N2209/046 ,  H04N9/88

摘要： In an image processing system, an outline correction of image data and a smoothing for noise removal are performed. Generally, the outline correction is performed by emphasizing a high frequency component, and the smoothing is performed by removing the high frequency component. Within one image data, by properly switching the emphasizing and the removal of the high frequency component, while the outline correction is carried out, the noise removal can be performed. An absolute value of a high frequency component Fa extracted from the image data Pi by a first differential filter is compared with a discrimination threshold value K, and depending on the comparison result, either Fa or -Fa as the correction data Fo is selected by a selector and is added to the image data Pi. As a result, when Fo=Fa, the high frequency component is emphasized, and when Fo=-Fa, smoothing is performed. A pixel interpolation and a multi picture formation are also disclosed.

摘要翻译： 在图像处理系统中，执行图像数据的轮廓校正和噪声去除的平滑处理。 通常，通过强调高频分量来执行轮廓校正，并且通过去除高频分量来执行平滑。 在一个图像数据内，通过适当地切换高频分量的强调和去除，在执行轮廓校正的同时，可以执行噪声去除。 将通过第一差分滤波器从图像数据Pi提取的高频分量Fa的绝对值与鉴别阈值K进行比较，并且根据比较结果，作为校正数据Fo的Fa或-Fa被选择为 选择器，并被添加到图像数据Pi。 结果，当Fo = Fa时，强调高频分量，并且当Fo = -Fa时，进行平滑处理。 还公开了像素内插和多图像形成。

公开(公告)号：US5550936A

公开(公告)日：1996-08-27

申请号：US238946

申请日：1994-05-06

申请人： Jun Someya ,  Kazuhiro Chiba ,  Kazuyoshi Watabu ,  Yoshie Yamamoto ,  Mamoru Inamura ,  Takeshi Nagayama ,  Noriko Bamba ,  Noboru Sunada

发明人： Jun Someya ,  Kazuhiro Chiba ,  Kazuyoshi Watabu ,  Yoshie Yamamoto ,  Mamoru Inamura ,  Takeshi Nagayama ,  Noriko Bamba ,  Noboru Sunada

IPC分类号： G06T3/40 ,  G06T5/00 ,  G06T5/20 ,  H04N1/409 ,  H04N5/208 ,  H04N5/21 ,  H04N5/367 ,  H04N9/04 ,  G06K9/00

CPC分类号： G06T3/4007 ,  G06T5/004 ,  G06T5/20 ,  H04N1/409 ,  H04N5/208 ,  H04N5/21 ,  H04N5/367 ,  H04N9/045 ,  H04N2209/041 ,  H04N2209/046 ,  H04N9/88

摘要： In an image processing system, an outline correction of image data and a smoothing for noise removal are performed. Generally, the outline correction is performed by emphasizing a high frequency component, and the smoothing is performed by removing the high frequency component. Within one image data, by properly switching the emphasizing and the removal of the high frequency component, while the outline correction is carried out, the noise removal can be performed. An absolute value of a high frequency component Fa extracted from the image data Pi by a first differential filter is compared with a discrimination threshold value K, and depending on the comparison result, either Fa or -Fa as the correction data Fo is selected by a selector and is added to the image data Pi. As a result, when Fo=Fa, the high frequency component is emphasized, and when Fo=-Fa, smoothing is performed. A pixel interpolation and a multi picture formation are also disclosed.

摘要翻译： 在图像处理系统中，执行图像数据的轮廓校正和噪声去除的平滑处理。 通常，通过强调高频分量来执行轮廓校正，并且通过去除高频分量来执行平滑。 在一个图像数据内，通过适当地切换高频分量的强调和去除，在执行轮廓校正的同时，可以执行噪声去除。 将通过第一差分滤波器从图像数据Pi提取的高频分量Fa的绝对值与鉴别阈值K进行比较，并且根据比较结果，作为校正数据Fo的Fa或-Fa被选择为 选择器，并被添加到图像数据Pi。 结果，当Fo = Fa时，强调高频分量，并且当Fo = -Fa时，进行平滑处理。 还公开了像素内插和多图像形成。