System and method for dithering with reduced memory

• US 7,236,269 B2
• Filed: 11/05/2002
• Issued: 06/26/2007
• Est. Priority Date: 11/05/2001
• Status: Expired due to Fees

• Current Assignee: Chrontel, Inc.
• Original Assignee: Chrontel, Inc.
• Inventors: Donovan, Timothy J., Nguyen, Dung, Long, Meng
• Primary Examiner(s): Lee; Thomas D.
• Assistant Examiner(s): BRINICH, STEPHEN M
• Application Number:
  US10/288,781
• Publication Number:
  US 20030122847A1
• Time in Patent Office: 1,694 Days
• Field of Search: 358 321- 322, 358 303- 306, 358 314- 315, 358/3.01, 358/1.9, 358/2.1
• US Class Current: 358/3.22
• CPC Class Codes: G09G 3/2059 using error diffusion

Abstract

A dithering system yielding two-dimensional dither functioning is implemented without line memories. For each primary input color, a feedback loop outputs an color input signal plus error that can be preset to different values. The desired result is that vertical artifacts on a display formed from the output signals are relocated to different locations on consecutive display lines. If signal magnitude from the feedback loop output exceeds the magnitude of the video system creating the display, signal magnitude is preset to a value representing error at the start of the display line.

Claims

1. A method to dither an image representable by lines of clockable 24-bit pixels, each of said pixels represented as red, green, and blue primary color signals using 8-bits of information, the method comprising the following steps:

(a) for each of said 8-bit wide red, green, and blue color signals, defining an error signal based upon less than 8-bit width color representation; and

(b) for each of said red, green, and blue color signals, presetting said error signal to different values for different said lines comprising said image; and

(c) generating said image lines so that a vertical artifact in an image formed from said colors using said different preset error for said different lines will appear at a different location on at least a plurality of adjacent lines of said image.

2. The method of claim 1, further including:

if magnitude of any of said 8-bit wide color signals plus said error at step (b) exceeds a threshold associated with a system displaying said image, presetting said error to magnitude of said error at the start of a line of said image;

wherein vertical artifact in an image formed from said colors is reduced even if said magnitude at step (b) is large.

3. The method of claim 1, further including:

(d) for each of said colors, feeding back said error signal defined at step (a) to generate a sum of said 8-bit wide color and said error signal;

(e) for each of said red, green, and blue colors, bit-limiting said sum generated at step (d); and

(f) outputting as red, green, and blue color data bit-limited sums created at step (e);

wherein an image formed from color data output at step (f) exhibits reduced quantization error.

4. The method of claim 1, wherein for each of said red, green, and blue colors, at step (a) said less than 8-bit width color representation is 6-bit width color representation.

5. The method of claim 1, wherein for each of said red, green, and blue colors, at step (a), said less than 8-bit width color representation is 6-bit width color representation, and said error signal is defined as two least significant bits (LSB) of said 8-bit width color representation.

6. The method of claim 3, wherein at step (d) each of said sums is limited in magnitude to 2⁸−

1.

7. The method of claim 3, wherein at step (f) each of said bit-limited sums is limited to 6-bit width.

8. The method of claim 2, wherein said threshold has magnitude 252.

9. A method to dither an image representable by clockable 8-bit wide red, green, and blue color signals, the method comprising the following steps:

(a) for each of said 8-bit wide red, green, and blue color signals generating a one-dimensional dither function having less than 8-bit color width;

(b) for each 8-bit wide red, green, and blue color signals, using feedback to generate an error signal based upon bit-limiting said color signals to less than 8-bit width;

(c) if magnitude of one of said 8-bit wide color signals plus said error at step (b) exceeds a threshold associated with a system displaying said image, presetting said error to magnitude of said error at the start of a line of said image; and

(d) generating a plurality of adjacent lines of said image to realize a two-dimensional dither function in which any vertical artifacts in the image formed from said colors is reduced even if said magnitude at step (c) is large.

10. The method of claim 9, wherein at step (b), for each of said red, green, and blue color signals, said less than 8-bit width is 6-bit width.

11. A system to dither an image representable by lines of clockable 24-bit pixels, each of said pixels represented as red, green, and blue primary color signals using 8-bits of information, the system comprising:

for each of said 8-bit wide red, green, and blue color signals, means for generating an output signal comprising one of said 8-bit wide color signals plus an error signal based upon less than 8-bit wide color representation;

means for presetting magnitude of each said output signal to different values for different said lines comprising said image; and

means for generating said image lines so that a vertical artifact in an image formed from said colors using said different preset error for said different lines will appear at a different location on at least a plurality of adjacent lines of said image.

12. The system of claim 11, further including:

means for presetting said magnitude of a set output signal to a magnitude representing output at the start of a line of said image when said output signal magnitude exceeds a threshold associated with a system displaying said image;

wherein vertical artifact in an image formed from said colors is reduced even if magnitude of said error is large.

13. The system of claim 11, further including:

a feedback sub-system for each of said red, green, and blue colors in which said error signal is fed-back to generate a sum of said 8-bit wide colors and said error signal;

for each of said red, green, and blue pixels, a bit-limiter coupled to limit each said sum generated by said feedback sub-system;

wherein an image formed from color data output from each said bit-limiter exhibits reduced quantization error.

14. The system of claim 11, wherein for each of said red, green, and blue colors, said less than 8-bit wide color representation is 6-bit width.

15. The system of claim 11, wherein for each of said red, green, and blue colors, said means for generating an output signal defines said output signal as two least significant bits (LSB) of said 8-bit width color representation.

16. The system of claim 13, wherein said bit limiter limits each of said sums in magnitude to 2⁸−

1.

17. The system of claim 13, wherein said bit limiter limits color data output to 6-bit width.

18. The system of claim 12, wherein said threshold has magnitude 252.

19. A system to dither an image representable by lines of clockable 24-bit pixels, each of said pixels represented as red, green, and blue primary color signals using 8-bits of information, the system comprising:

for each of said 8-bit wide red, green, and blue color signals, means for generating an error signal based upon less than 8-bit wide color representation;

for each of said 8-bit wide red, green, and blue color signals, means for presetting said error signal to different values for different said lines comprising said image;

means for presetting said error signal at a start of a line of said image to a magnitude representing error if any of said 8-bit wide color signals plus said error exceeds a threshold magnitude associated with said system; and

means for generating said image lines so that a wherein vertical artifact in an image formed from said colors using said different preset error for said different lines will appear at a different location on at least a plurality of adjacent lines of said image.

20. The system of claim 19, further including:

a feedback sub-system for each of said red, green, and blue color signals, in which said error signal is fed-back to generate a sum of an 8-bit wide color signal and said error signal; and

for each of said red, green, and blue pixels, a bit-limiter coupled to limits each said sum generated by said feedback sub-system.

21. A method for dithering at least a portion of an image representable by lines of clockable 24-bit pixels, each of said pixels represented as red, green, and blue primary color signals using 8-bits of information, the method comprising the steps:

(a) for each of said 8-bit wide red, green, and blue color signals, defining an error signal based upon less than 8-bit width color representation;

(b) for each of said red, green, and blue color signals, presetting said error signal to different values for different said lines comprising said image;

(c) generating plurality of adjacent lines so that any vertical artifacts in an image formed from said colors that occur in said plurality of adjacent lines will appear at a different location in said at least two adjacent lines of said image;

(d) if magnitude of any of said 8-bit wide color signals plus said error at step (b) exceeds a threshold associated with a system displaying said image, presetting said error to magnitude of said error at the start of a line of said image so that any vertical artifact in an image formed from said colors is reduced even if said magnitude at step (b) is large;

(e) for each of said colors, feeding back said error signal defined at step (a) to generate a sum of said 8-bit wide color and said error signal;

(f) for each of said red, green, and blue colors, bit-limiting said sum generated at step (c); and

(g) outputting as red, green, and blue color data bit-limited sums created at step (f) so that an image formed from color data output at step (e) exhibits reduced quantization error.

22. A system to dither an image representable by lines of clockable 24-bit pixels, each of said pixels represented as red, green, and blue primary color signals using 8-bits of information, the system comprising:

for each of said 8-bit wide red, green, and blue color signals, means for generating an error signal based upon less than 8-bit wide color representation;

for each of said 8-bit wide red, green, and blue color signals, means for presetting said error signal to different values for different said lines comprising said image;

at least one circuit for presetting said error signal to a magnitude representing error at a start of a line of said image if any of said 8-bit wide color signals plus said error exceeds a threshold magnitude associated with said system; and

at least one circuit for generating a plurality of adjacent lines of said image using said different values of preset magnitude in said different lines so that any vertical artifacts in an image formed from said colors signals that occur in said plurality of adjacent lines will appear at a different location in said plurality of adjacent lines of said image.