| 摘要 |
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. |
| 主权项 |
1. A method comprising:
converting an original image into an input image having a lower bit count per pixel, said input image having pixel error based on said lower bit count per pixel, said converting diffusing said pixel error of each pixel of said input image to adjacent pixels using an error weighting matrix, said input image and consuming less storage and data processing resources relative to said original image; receiving said input image represented in an input image matrix into an image path of a printing device; and processing said input image through said image path to produce an output image from said input image, said image path having storage and data processing constraints that prevent said image path from processing said original image, said storage and data processing constraints allow said image path to process said input image, and said processing comprising:
determining, using said error weighting matrix, for each pixel in said input image, an amount of error added based on said adjacent pixels during said converting by processing each pixel individually, starting with a last matrix location at the end of said input image matrix and progressing back, pixel-by-pixel, to a first matrix location at the beginning of said input image matrix;removing, from each pixel in said input image, said amount of error added based on said adjacent pixels by processing each pixel individually, starting with said last matrix location at the end of said input image matrix and progressing back, pixel-by-pixel, to said first matrix location at the beginning of said input image matrix; andreconstructing said input image using a filter after said removing said amount of error added based on said adjacent pixels from all pixels in said input image to produce said output image, said filter comprising a low-pass matrix filter, said reconstructing:
using said low-pass matrix filter to multiply numerical values of said low-pass matrix filter by corresponding underlying pixel values of pixels adjacent a center pixel over which said low-pass matrix filter is centered to produce products;adding said products together to produce a sum; andmultiplying said sum by a reciprocal of a total of said low-pass matrix filter values to produce a resulting value of said center pixel, and said reconstructing producing pixels having a higher bit count per pixel relative to said lower bit count per pixel of said input image. |
