Granted Patent B2

US 9,007,395 · App. 12/615,594 · Granted Apr 14, 2015

Abstract

Claims (29)

1. An apparatus, comprising:

mapping logic configured to produce a high-resolution pixel from a low-resolution pixel;

contour segmentation logic configured to classify the high-resolution pixel as being in a smooth region of an image based on at least one of a gradient value and a variance value associated with the low-resolution pixel;

contour refinement logic configured to:

identify a set of neighboring high-resolution pixels with respect to the high-resolution pixel; and

selectively re-classify the high-resolution pixel as not being in the smooth region of the image when any pixels in the set of neighboring high-resolution pixels is classified as not being in the smooth region of the image; and

range adjustment logic configured to selectively filter the high-resolution pixel based on whether the high-resolution pixel remains classified as being in the smooth region of the image,

where the mapping logic is configured to produce the high-resolution pixel from the low-resolution pixel by operating on an individual color channel in a color space associated with the low-resolution pixel, and

where the range adjustment logic is configured to selectively filter the high-resolution pixel in a different color channel of the color space.

2. The apparatus of claim 1 , where the mapping logic is further configured to produce the high-resolution pixel from the low-resolution pixel according to a one-to-one mapping.

3. The apparatus of claim 1 , where the mapping logic is further configured to produce the high-resolution pixel from the low-resolution pixel by performing one or more of, bit appending, zero padding, adaptive noise processing, quantizing processing, conversion processing, consulting a lookup table, employing a hash function, dithering, or half-toning.

4. The apparatus of claim 1 , where the contour segmentation logic is further configured to calculate the gradient value based on change in pixel values in a line of pixels including the low-resolution pixel, where the line of pixels is confined to a neighborhood of pixels in the image.

5. The apparatus of claim 1 , where the contour segmentation logic is further configured to calculate the variance value based on a difference amount between the low-resolution pixel and a set of related low-resolution pixels, where the set of related low-resolution pixels is confined to a neighborhood of pixels in the image.

6. The apparatus of claim 1 , where the contour segmentation logic is further configured to classify the high-resolution pixel as being in the smooth region of the image upon determining that the gradient value is less than a gradient threshold and that the variance value is less than a variance threshold.

7. The apparatus of claim 6 , where at least one of the gradient threshold and the variance threshold are user configurable.

8. The apparatus of claim 1 , where the mapping logic is further configured to operate on an individual color channel associated with the low-resolution pixel.

9. The apparatus of claim 1 , where the contour refinement logic is further configured to classify the high-resolution pixel as not being in the smooth region of the image as a function of at least one of the set of neighboring high-resolution pixels being classified as not being in the smooth region.

10. The apparatus of claim 1 , where the range adjustment logic includes one or more of an adaptive low pass filter and an adaptive averaging filter that computes a filter value as a function of the average of related high-resolution pixels located in a line that includes the high-resolution pixel.

11. The apparatus of claim 1 , where the mapping logic is configured to produce, on an intra-frame basis, a set of high-resolution pixels associated with a video stream from a set of low-resolution pixels.

12. A method, comprising:

enhancing a bit depth representation of image content from a lower bit depth representation to a higher bit depth representation;

initially classifying a pixel having the higher bit depth representation that is susceptible to producing a contouring artifact;

identifying a neighboring set of pixels having the higher bit depth representation associated with the initially classified pixel;

re-classifying the pixel as not being susceptible to producing the contouring artifact upon determining that one or more pixels in the neighboring set of pixels is not susceptible to producing the contouring artifact;

selectively filtering the pixel when the pixel remains classified as being susceptible to producing the contouring artifact to mitigate the contouring artifact,

where re-classifying the pixel is based on luma channel data, and where selectively filtering the pixel is based on chroma channel data.

13. The method of claim 12 , where enhancing the bit depth representation comprises one or more of, zero padding a pixel having the lower bit depth representation, performing adaptive noise processing on the pixel having the lower bit depth representation, acquiring a value from a lookup table, acquiring a value from a hash function, acquiring a value from a dithering function, and acquiring a value from a half-toning function.

14. The method of claim 12 , where initially classifying the pixel having the higher bit depth representation comprises initially classifying the pixel having the higher bit depth representation as a function of gradient values and variance values associated with the pixel.

15. The method of claim 12 , where selectively filtering the pixel comprises applying one or more of an adaptive low pass filter or an adaptive averaging filter to the pixel.

Assignments (9)