Image processing method and apparatus for 3D video
First Claim
Patent Images
1. An image processing method performed by a computer, the method comprising:
- identifying a first depth image;
upsampling the first depth image to a second depth image having an increased number of pixels compared to the first depth image; and
expanding a foreground area of the upsampled second depth image by applying a dilation filter to the upsampled second depth image,wherein the upsampling is performed by interpolating the first depth image,wherein the expanding of the foreground area of the upsampled second depth image comprises expanding the foreground area by applying a dilation filter having a size that is adjusted based on characteristic information of the first depth image,wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, andwherein a size of the dilation filter is determined based on a resolution of the first depth image.
1 Assignment
0 Petitions
Accused Products
Abstract
An image processing method and apparatus for three-dimensional (3D) video is provided. The image processing method may include identifying a first depth image, and upsampling the first depth image to a second depth image having expanded foreground area than the first depth image.
14 Citations
29 Claims
-
1. An image processing method performed by a computer, the method comprising:
-
identifying a first depth image; upsampling the first depth image to a second depth image having an increased number of pixels compared to the first depth image; and expanding a foreground area of the upsampled second depth image by applying a dilation filter to the upsampled second depth image, wherein the upsampling is performed by interpolating the first depth image, wherein the expanding of the foreground area of the upsampled second depth image comprises expanding the foreground area by applying a dilation filter having a size that is adjusted based on characteristic information of the first depth image, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image.
-
-
2. An image processing method performed by a computer, the method comprising:
-
identifying a first depth image; upsampling the first depth image to a second depth image by interpolating the first depth image; and expanding a foreground area of the second depth image, wherein the expanding of the foreground area of the second depth image comprises expanding the foreground area of the second depth image by applying a dilation filter to the second depth image, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (3, 4)
-
-
5. An image processing method performed by a computer, the method comprising:
-
outputting a second depth image by resampling a first depth image; predicting a color image using the second depth image; encoding the color image based on the prediction; and encoding a size of a dilation filter used in resampling the first depth image; wherein the outputting of the second depth image by resampling the first depth image comprises; upsampling the first depth image to the second depth image; and expanding a foreground area of the second depth image using a dilation filter, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (6)
-
-
7. An image processing method performed by a computer, the method comprising:
-
outputting a second depth image by resampling a first depth image extracted from a bit stream; predicting a color image extracted from the bit stream, using the second depth image; and decoding the color image based on the prediction, wherein the outputting of the second depth image by resampling the first depth image comprises; upsampling the first depth image to the second depth image; and expanding a foreground area of the second depth image using a dilation filter, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (8, 9)
-
-
10. An image processing method performed by a computer, the method comprising:
-
identifying a decoded depth image; and resampling the depth image to a resolution of a color image, through post-processing, when a resolution of the depth image is different from the resolution of the color image, wherein the resampling the depth image comprises; upsampling the depth image to the resolution of the color image; and expanding a foreground area of the depth image using a dilation filter, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the depth image.
-
-
11. An image processing method performed by a computer, the method comprising:
-
warping a color image by referring to a depth image; performing a view synthesis prediction based on the warping result of the color image; and encoding the color image based on the view synthesis prediction, wherein the warping of the color image comprises selecting a maximum pixel value among pixel values of depth images of adjacent views for pixels of the color image.
-
-
12. An image processing method performed by a computer, the method comprising:
-
warping a compressed color image by referring to a first depth image reconstructed from a bit stream; performing a view synthesis prediction based on the warping result of the color image; and decoding the color image based on the view synthesis prediction, wherein the warping of the color image comprises selecting a maximum pixel value among pixel values of depth images of adjacent views for pixels of the color image.
-
-
13. An image processing method performed by a computer, the method comprising:
-
identifying a first depth image; upsampling the first depth image to a second depth image having an increased number of pixels compared to the first depth image, wherein the upsampling is performed by interpolating the first depth image; and outputting the second depth image having an expanded foreground area compared with the first depth image by applying a dilation filter to the upsampled second depth image, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image.
-
-
14. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; an image identifying unit to identify a first depth image; and an upsampling unit to upsample the first depth image to a second depth image having an increased number of pixels compared to the first depth image, wherein the upsampling is performed by interpolating the first depth image and applying a dilation filter to the interpolated first depth image, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image.
-
-
15. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; an image identifying unit to identify a first depth image; an upsampling unit to upsample the first depth image to a second depth image by interpolating the first depth image; and an area expanding unit to expand a foreground area of the upsampled second depth image using a dilation filter, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (16)
-
-
17. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; a resampling unit to output an upsampled second depth image by resampling a downsampled first depth image; a predicting unit to predict a color image using the second depth image; and an image encoding unit to encode the color image based on the prediction, wherein the resampling unit upsamples the first depth image to the second depth image; and
expands a foreground area of the second depth image using a dilation filter,wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (18)
-
-
19. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; a resampling unit to output an upsampled second depth image by resampling a downsampled first depth image included in a bit stream; a predicting unit to predict a color image included in the bit stream using the second depth image; and an image encoding unit to decode the color image based on the prediction, wherein the resampling unit upsamples the first depth image to the second depth image; and
expands a foreground area of the second depth image using a dilation filter,wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image.
-
-
20. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; a predicting unit to warp a color image by referring to a first depth image and to perform a view synthesis prediction based on the warping result of the color image; and an image encoding unit to encode the color image based on the view synthesis prediction, wherein the predicting unit performs warping the color image by selecting a maximum pixel value among pixel values of depth images of adjacent views for pixels of the color image.
-
-
21. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; a predicting unit to warp a compressed color image by referring to a first depth image reconstructed from a bit stream and to perform a view synthesis prediction based on the warping result of the color image; and an image decoding unit to decode the color image based on the view synthesis prediction, wherein the predicting unit performs warping the color image by selecting a maximum pixel value among pixel values of depth images of adjacent views for pixels of the color image.
-
-
22. An image processing apparatus, comprising:
-
a processor to control one or more processor-executable units; an image identifying unit to identify a first depth image; an upsampling unit to upsample the first depth image to a second depth image having an increased number of pixels compared to the first depth image, wherein the upsampling is performed by interpolating the first depth image; and an image output unit to output the second depth image having an expanded foreground area compared with the first depth image using a dilation filter, wherein the first depth image has a resolution lower than or equal to the second depth image, wherein the dilation filter determines a pixel value of a new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the first depth image. - View Dependent Claims (23)
-
-
24. A method of adjusting a depth image, the method comprising:
-
identifying a depth image having a lower resolution than a corresponding color image; upsampling the depth image by interpolating the depth image to increase the resolution of the depth image when the depth image is determined as having a lower resolution than the corresponding color image; and expanding, by way of a processor, a foreground area of the upsampled depth image by applying a dilation filter to the upsampled depth image, wherein the dilation filter determines a pixel value of new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the depth image. - View Dependent Claims (25, 26, 27)
-
-
28. A 3-dimensional display including an apparatus for generating multi-views, the 3-dimensional display comprising:
-
a processor to control one or more processor-executable units; an image identifying unit to determine a depth image having a lower resolution than a corresponding color image; and an upsampling unit to upsample the depth image for increasing the resolution of the depth image when the depth image is determined as having a lower resolution than the corresponding color image, and to expand a foreground area of the upsampled depth image by applying a dilation filter to the upsampled depth image, wherein the dilation filter determines a pixel value of new pixel generated by upsampling, using a highest pixel value among a pixel value corresponding to neighbor pixels, and wherein a size of the dilation filter is determined based on a resolution of the depth image. - View Dependent Claims (29)
-
Specification