Error diffusion with output and input based feedback

• US 5,325,211 A
• Filed: 01/04/1993
• Issued: 06/28/1994
• Est. Priority Date: 01/04/1993
• Status: Expired due to Fees

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Eschbach, Reiner
• Primary Examiner(s): Brinich, Stephen
• Assistant Examiner(s): LEE, CHEUKFAN
• Application Number:
  US08/000,166
• Time in Patent Office: 540 Days
• Field of Search: 358/447, 358/455, 358/456, 358/466, 358/448, 358/443, 358/462, 358/445, 358/468, 358/467, 358/444, 358/453, 382/50, 382/54, 355/214
• US Class Current: 358/466
• CPC Class Codes: H04N 1/4053 with threshold modulated relative to input image data or vice versa

Abstract

A method of quantizing pixels in an image formed by a plurality of pixels, each pixel representing an optical density of the image at a location therewith, and having an original optical density value selected from one of a set of `c` original optical density values that has a number of members larger than a desired output set of `d` optical density values, through a process of error diffusion. In this process, `c` and `d` are integer values representing pixel depth. For an input image so comprised, initially, an image signal is initially modified in accordance with previously determined error. Thereafter, during the quantization step of the process, the threshold of an error diffusion process is dynamically adjusted in accordance with the original image to selectively control an amount of edge enhancement introduced into the image. The threshold level is selectively modified on a pixel by pixel basis and may be used to increase or decrease the edge enhancement of the output digital image, thus, more closely representing the original detail and edge sharpness of the continuous tone input image. The threshold of the error diffusion process is also varied responsive to the image that is output, by providing a feedback response to the thresholding process, to vary the threshold level to favor a determination of an output pixel having the same state as an adjacent pixel. Based on the dynamically varying threshold, an output pixel is determined, and the error or difference between the output and the modified image signal is determined. This error is added on a weighted basis to a predetermined group of neighboring image signals, in accordance with the image signal modification initially described.

Claims

1. A document processing system for preparing a document derived originally with a device describing each signal in terms of c levels, for printing at a printer responding to printers signals having d levels, while maintaining gray density with the image, comprising:

a scanner, producing a set of input signals representing optical density at discrete points within a document, said input signals having magnitudes varying over c levels;

means for adding to each input signal a fractional portion of an error signal of at least one previously processed input signal;

means for comparing each modified input signal with at least one reference signal to determine which of d printer output signals best represents the modified input signal;

a printer, responsive to said printer output signals reproduce the document in accordance therewith;

means for determining the difference between the determined printer output signal and the modified input signal, and generating at least one error signal responsive indicative thereof;

means for varying the reference signal as a function of the input signal and at least one previous printer output signal.

2. The device as defined in claim 1, wherein said input signals derived by said scanner are N-bit digital signals representing gray at a number of levels equal to 2^N.

3. The device as defined in claim 1, wherein said printer output signals directed to said printer are one bit digital signals representing either black or white.

4. The device as defined in claim 1, wherein said means for varying the reference signal as a function of the input signal and at least one previous printer output signal includes:

means for generating an input threshold signal;

a plurality of registers storing printer output signals for previously processed input signals at preselected positions with respect to a current input signal;

means for determining a weighted fraction of each the stored printer output signal;

means for summing the weighted fractions of the stored printer output signals;

means for adding the input threshold signal and the weighted sum of the stored printer output signals to derive the reference signal.

5. The device as defined in claim 1, wherein said means for determining the difference between the determined printer signal and the modified input signal, and generating at least one error signal responsive indicative thereof includes:

means for deriving error signals representing the difference between the determined printer output signal and the modified input signal;

a plurality of registers storing said error signals for previously processed input signals at preselected positions with respect to a current input signal;

means for determining a weighted fraction of each of the stored error signals;

means for summing the weighted fractions of the stored error signals;

means for adding the sum of weighted fractions of the stored error signals to the input signal, prior to comparison with said reference signal.

6. A document processing system for preparing a document derived originally with a device describing each signal in terms of c levels, for display at a display responding to display signals having d levels, while maintaining gray density within the image, comprising:

a scanner, producing a set of input signals representing optical density at discrete points within a document, said input signals having magnitudes varying over c levels;

means for adding to each input signal a fractional portion of an error signal of at least one previously processed input signal;

means for comparing each modified input signal with at least one reference signal to determine which of d display output signals best represents the modified input signal;

a display, responsive to said display output signals to reproduce the document in accordance therewith;

means for determining the difference between the determined display output signal and the modified input signal, and generating at least one error signal responsive indicative thereof;

means for varying the reference signal as a function of the input signal and at least one previous display output signal.

7. The device as defined in claim 6, wherein said input signals derived by said scanner are N-bit digital signals representing gray at a number of levels equal to 2^N.

8. The device as defined in claim 7, wherein said display output signals directed to said display are one bit digital signals representing either black or white.

9. The device as defined in claim 6, wherein said means for varying the reference signal as a function of the input signal and at least one previous display output signal includes:

means for generating an input threshold signal;

a plurality of registers storing display output signals for previously processed input signals at preselected positions with respect to a current input signal;

means for multiplying each stored display output signals by a predetermined hysteresis function;

means for summing the outputs of the multiplying means to derive an output based feedback signal;

means for adding the input threshold signal and the output based feedback signal to derive the reference signal.

10. The device as defined in claim 6, wherein said means for determining the difference between the determined display signal and the modified input signal, and generating at least one error signal responsive indicative thereof includes:

means for deriving error signals representing the difference between the determined display output signal and the modified input signal;

a plurality of registers storing said error signals for previously processed input signals at preselected positions with respect to a current input signal;

means for determining a weighted fraction of each of the stored error signals;

means for summing the weighted fractions of the stored error signals;

means for adding the sum of weighted fractions of the stored error signals to the input signal, prior to comparison with said reference signal.

11. A document processing system for preparing a document derived originally with a device describing each signal in terms of c levels, for printing at a printer responding to printer signals having d levels, while maintaining gray density with the image, comprising:

a computer workstation, operating in accordance with document production software, producing a set of input signals representing optical density at discrete points within a document, said input signals having magnitudes varying over c levels;

means for adding to each input signal a fractional portion of an error signal of at least one previously processed input signal;

means for comparing each modified input signal with at least one reference signal to determine which of d printer output signals best represents the modified input signal;

a printer, responsive to said printer output signals to reproduce the document in accordance therewith;

means for determining the difference between the determined printer output signal and the modified input signal, and generating at least one error signal responsive indicative thereof;

means for varying the reference signal as a function of the input signal and at least one previous printer output signal.

12. The device as defined in claim 11, wherein said input signals derived by said computer workstation are N-bit digital signals representing gray at a number of levels equal to 2^N.

13. The device as defined in claim 12, wherein said printer output signals directed to said printer are one bit digital signals representing either black or white.

14. The device as defined in claim 11, wherein said means for varying the reference signal as a function of the input signal and at least one previous printer output signal includes:

means for generating an input threshold signal;

a plurality of registers storing printer output signals for previously processed input signals at preselected positions with respect to a current input signal;

means for multiplying each stored printer output signals by a predetermined hysteresis function;

means for summing the outputs of the multiplying means to derive an output based feedback signal;

means for adding the input threshold signal and the output based feedback signal to derive the reference signal.

15. The device as defined in claim 11, wherein said means for determining the difference between the determined printer output signal and the modified input signal, and generating at least one error signal responsive indicative thereof includes:

means for deriving error signals representing the difference between the determined printer output signal and the modified input signal;

a plurality of registers storing said error signals for previously processed input signals at preselected positions with respect to a current input signal;

means for determining a weighted fraction of each of the stored error signals;

means for summing the weighted fractions of the stored error signals;

means for adding the sum of weighted fractions of the stored error signals to the input signal, prior to comparison with said reference signal.

16. A method of quantizing pixels in an image formed by a plurality of pixels, each pixel representing an optical density of the image at a location therewithin as determined by an gray pixel producing device, and having an original optical density value selected from one of a set of `c` original optical density values that has a number of members larger than a desired printer set of `d` optical density values required of an output device, the steps including

receiving an image signal from a said gray pixel producing devices;

modifying each gray pixel in accordance with previously determined local error;

setting a threshold level as a function of a current gray pixel and a number of previously pixels previously directed to said output device;

thresholding each gray pixel with the threshold level set, to derive black or white pixels, suitable to drive the output device;

determining a gray difference between the black or white output device driving pixels and the modified gray pixels, and generating an error signal representative of the gray difference;

storing said error signals, for subsequent formation of a local error and modification of the image signal modification.

17. A document processing system for preparing a document derived originally with a device describing each signal in terms of c levels, for printing at a printer responding to printer signals having d levels, while maintaining gray density with the image, comprising:

a scanner, producing a set of input signals representing optical density at discrete points within a document, said input signals having magnitudes varying over c levels;

means for adding to each input signal a fractional portion of an error signal of at least one previously processed input signal;

means for comparing each modified input signal with at least one reference signal to determine which of d printer output signals best represents the modified input signal;

a printer, responsive to said printer output signals reproduce the document in accordance therewith;

means for determining the difference between the determined printer output signal and the modified input signal, and generating at least one error signal responsive indicative thereof;

means for varying the reference signal as a function of the input signal and at least one previous printer output signal including;

first input based feedback means, producing a first signal dynamically varied as a function of the input signals which tends to produce a reference signal enhancing image edges;

second output based feedback means, producing a second signal varied in accordance with a number of neighboring printer output signals which have a same level, in order to cluster printer output signals for improved reproducibility, but tending to dampen edge response;

combining means for combining said first and second signals to generate a varying reference signal.