Image processor and method for processing pixel data

• US 5,301,038 A
• Filed: 06/05/1992
• Issued: 04/05/1994
• Est. Priority Date: 03/07/1990
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Todd, Stephen J. P.
• Primary Examiner(s): Coles, Sr., Edward L.
• Assistant Examiner(s): Rogers, Scott A.
• Application Number:
  US07/895,182
• Time in Patent Office: 669 Days
• Field of Search: 358/465, 358/466, 358/455, 358/458, 358/456, 358/448, 358/463, 358/75, 358/80, 358/530, 358/532, 358/534, 358/462, 358/445, 358/447, 358/443, 382/50, 382/52, 382/53, 382/54, 395/128
• US Class Current: 382/269
• CPC Class Codes: G06T 2200/12 involving antialiasing
• CPC Class Codes: G06T 5/70 Denoising; Smoothing

Abstract

An image processor has image processing logic for detecting pixel data representing stepped chrominance or luminance transitions characteristic of staircase aliasing in a source image and for producing an output image including selective antialiasing on the detected stepped transitions. The image processing logic operates as a post processor, but is able to give much better results than conventional convolution filtering.

Claims

1. An image processor for processing pixel data defining an image, said pixel data defining an intensity for each of a plurality of pixels having a plurality of gradiations between a predetermined minimum and maximum, said image processor including image processing logic comprising:

means for determining the intensity difference between adjacent pixels of said image;

means for comparing said intensity difference with a predetermined nonzero threshold that is less than the difference between said predetermined minimum and said predetermined maximum to detect stepped intensity transitions characteristic of staircase aliasing in the image;

means for determining a stepped intensity range indicating the range of pixels through which said intensity remains essentially constant;

means for determining an antialiased pixel value for each of said range of pixels, said antialiased pixel value being a function of said range and said adjacent pixel intensities; and

means for producing an output image including said antialiased pixel values on the detected stepped transitions.

2. An image processor as claimed in claim 1 wherein the image processing logic comprises staircase point detection logic for detecting a staircase point at which there is a step in an intensity transition from one side of a line of pixels to the opposing side of the line of pixels, where the line of pixels is a row or a column of image pixels.

3. An image processor as claimed in claim 2 wherein the image processing logic comprises staircase segment logic for finding the far end of first and second staircase segments starting at the staircase point detected by the staircase point detection logic;

the staircase segment logic locating the end of a first staircase segment by testing for the position in a first direction along the line of pixels at which the transition ceases to be adjacent said opposing side of the line of pixels and testing for the position in the opposite direction along the line of pixels at which the transition ceases to be adjacent said one said of the line of pixels.

4. An image processor as claimed in claim 3 wherein the image processing logic comprises image modification logic for testing for predetermined conditions at the end of a staircase segment.

5. An image processor as claimed in claim 4 wherein the image modification logic modifies the pixel data in accordance with the predetermined condition detected for the staircase segment.

6. An image processor as claimed in claim 5 wherein the predetermined conditions tested for are:

(a) the transition steps across a further line of pixels;

(b) the transition steps back across the line of pixels; and

(c) neither (a) nor (b).

7. An image processor as claimed in claim 1 in which the image processing logic scans the image for times, in each scan a test being made for a different type of staircase point.

8. An image processor as claimed in claim 7 wherein the image processing logic performs a plurality of operations in parallel.

9. An image processor as claimed in claim 1 in which the image processing logic treats each pixel position in turn, for each pixel position testing for four different types of staircase point.

10. An image processor as claimed in claim 9 wherein the image processing logic performs a plurality of operations in parallel.

11. An image processor as claimed in claim 1 comprising source image storage for a source image and output image storage for the output image wherein the image processing logic performs an initial operation of copying the source image from the source image storage to the output image storage, the image processing logic then testing the source image in the source image storage to detect pixel data representing stepped intensity transitions characteristic of staircase aliasing and modifying data accordingly in the output image storage to produce the output image including antialiasing on the detected stepped transitions.

12. A data processing system comprising an image processor as claimed in claim 1.

13. An image processor as in claim 1 in which said pixel data defines an intensity for each of a plurality of colors for said pixels and in which intensity differences between adjacent pixels of said image in excess of a predetermined threshold for any of said colors are detected to detect said stepped transitions.

14. In a graphics system in which a two-dimensional image consisting of rows and columns of pixels is generated for display by a raster scan device, said image being represented by pixel data defining an intensity for each of said pixels having a plurality of gradiations between a predetermined minimum and maximum, a method for processing said pixel data comprising the steps of:

determining the intensity differences between adjacent pixels of said image;

comparing said intensity differences with a predetermined nonzero threshold that is less than the difference between said predetermined minimum and said predetermined maximum to detect stepped transitions characteristic of staircase aliasing in said image;

determining a stepped intensity range indicating the range of pixels through which said intensity remains essentially constant;

determining an antialiased pixel value for each of said range of pixels, said antialiased pixel value being a function of said range and said adjacent pixel intensities; and

modifying said pixel data for pixels along said detected transitions to product an output image including said antialiased pixel values along said transitions.

15. A method as in claim 14 in which each of said transitions has a segment extending along a line of pixels between a first end and a second end, said modifying step comprising the steps of:

testing for a predetermined condition at one end of a segment; and

modifying said pixel data for pixels along said segment in accordance with the predetermined condition detected at said one end of said segment.

16. A method as in claim 15 in which said one end of said segment is tested for the stepping of said transition across a further line of pixels.

17. A method as in claim 15 in which said one end of said segment is tested for the stepping of said transition back across said line of pixels.

18. A method as in claim 15 in which said one end of said segment is tested for the absence of a stepping of said transition back across said line of pixels or across a further line of pixels.

19. A method as in claim 14 in which said pixel data defines an intensity for each of a plurality of colors for said pixels and in which intensity differences between adjacent pixels of said image in excess of a predetermined threshold for any of said colors are detected to detect said stepped transitions.

20. A method as in claim 14 in which said pixel data is modified by blending intensity values for pixels on either said of said transitions.