公开(公告)号：US09521296B2

公开(公告)日：2016-12-13

申请号：US14469944

申请日：2014-08-27

Applicant: Xerox Corporation

Inventor： Dhevendra A. Palanivel ,  Sunil Kumar ,  Sainarayanan Gopalakrishnan ,  Xing Li

IPC: H04N1/405 ,  H04N1/40

CPC classification number: H04N1/4052 ,  H04N1/40075 ,  H04N2201/0094

Abstract: In systems and methods, an original image was previously converted to a lower bit count per pixel. The resulting image included pixel error based on the lower bit count per pixel. The conversion used an error weighting matrix (which diffuses the pixel error of each pixel to adjacent pixels). Methods and systems produce an output image from the lower bit count per pixel image using the error weighting matrix to determine, for each pixel, the amount of error added because of the adjacent pixels during the conversion process. Then, such methods and systems remove, from each pixel, the amount of error that was previously added based on the adjacent pixels. After removing the amount of error added based on the adjacent pixels from all pixels in the input image, an output image is reconstructed using a filter, such as a low pass filter.

Abstract translation: 在系统和方法中，原始图像先前被转换为每像素的较低位数。 所得到的图像包括基于每像素的较低位计数的像素误差。 转换使用误差加权矩阵（它将每个像素的像素误差扩散到相邻像素）。 方法和系统使用误差加权矩阵从每像素图像的较低位计数产生输出图像，以便为每个像素确定在转换过程期间由于相邻像素而添加的误差量。 然后，这些方法和系统从每个像素中移除先前根据相邻像素添加的错误量。 在从输入图像中的所有像素中除去基于相邻像素的误差量之后，使用诸如低通滤波器的滤波器重建输出图像。

公开(公告)号：US20160065786A1

公开(公告)日：2016-03-03

申请号：US14469944

申请日：2014-08-27

Applicant: Xerox Corporation

Inventor： Dhevendra A. Palanivel ,  Sunil Kumar ,  Sainarayanan Gopalakrishnan ,  Xing Li

IPC: H04N1/405

CPC classification number: H04N1/4052 ,  H04N1/40075 ,  H04N2201/0094

Abstract: In systems and methods, an original image was previously converted to a lower bit count per pixel. The resulting image included pixel error based on the lower bit count per pixel. The conversion used an error weighting matrix (which diffuses the pixel error of each pixel to adjacent pixels). Methods and systems produce an output image from the lower bit count per pixel image using the error weighting matrix to determine, for each pixel, the amount of error added because of the adjacent pixels during the conversion process. Then, such methods and systems remove, from each pixel, the amount of error that was previously added based on the adjacent pixels. After removing the amount of error added based on the adjacent pixels from all pixels in the input image, an output image is reconstructed using a filter, such as a low pass filter.

Abstract translation: 在系统和方法中，原始图像先前被转换为每像素的较低位数。 所得到的图像包括基于每像素的较低位计数的像素误差。 转换使用误差加权矩阵（它将每个像素的像素误差扩散到相邻像素）。 方法和系统使用误差加权矩阵从每像素图像的较低位计数产生输出图像，以为每个像素确定在转换过程期间由于相邻像素而添加的误差量。 然后，这些方法和系统从每个像素中移除先前根据相邻像素添加的错误量。 在从输入图像中的所有像素中除去基于相邻像素的误差量之后，使用诸如低通滤波器的滤波器重建输出图像。