Method and apparatus for encoding a point cloud representing three-dimensional objects
A colored three-dimensional point cloud is encoded in at least two depth and texture images. The proposed method and device use inverse project pixels of the first image and inverse project pixels to the second image. When the inverse projecting of a first pixel of the first image and the inverse projecting of a second pixel of the second image refer to a common generated point, that is two 3D points closer to each other than the imprecision due to the encoding and compression of the images, the color value of the pixel of first and second pixels having the lowest local depth variance is attributed to the common generated point.
1. A method of generating a three-dimensional colored point cloud, the method comprising:
obtaining a first image and a second image, the first and second images encoding a projection of an original three-dimensional colored point cloud, pixels of the first and second images comprising a depth attribute and a color attribute;
inverse projecting pixels of the first image and inverse projecting pixels of the second image; and
determining that the inverse projecting of a first pixel of the first image and the inverse projecting of a second pixel of the second image refer to a common generated point, and responsively attributing a color value to the common generated point by selecting a color value associated with one of the first pixel and the second pixel, wherein the associated pixel has a lowest local depth variance of the first pixel and the second pixel, wherein the local depth variance of a given pixel is computed for a region of n×n pixels around the given pixel, with n being an integer greater than or equal to three.
2. The method according to claim 1 , wherein the inverse projection of a pixel of an image is a three-dimension volume determined according to a resolution of the image in 3D space and to an imprecision value on the depth attribute of the pixel.
3. The method according to claim 2 , wherein determining that the first inverse projection of the first pixel and the second inverse projection of the second pixel refer to a common generated point comprises determining that a percentage of overlap of a volume of the first inverse projection and a volume of the second inverse projection is bigger than a threshold.
4. The method according to claim 2 , wherein determining that the first inverse projection of the first pixel and the second inverse projection of the second pixel refer to a common generated point comprises determining that a center of a volume of the first inverse projection is within a volume of the second inverse projection or that a center of the volume of the second inverse projection is within the volume of the first inverse projection.
5. The method according to claim 1 , wherein depth attributes and color attributes of at least one of the first and the second image are stored in a pair of one depth image and one texture image.
6. The method according to claim 1 , comprising decoding the first and second images from a received bitstream.
7. A device for generating a three-dimensional colored point cloud, the device comprising:
at least one processor; and
a memory, the memory having stored thereon instructions that are operative, when executed by the at least one processor, to cause the device to:
obtain a first image and a second image, the first and second images encoding a projection of an original three-dimensional colored point cloud, pixels of the first and second images comprising a depth attribute and a color attribute;
inverse project pixels of the first image and inverse project pixels of the second image; and
determine that the inverse projecting of a first pixel of the first image and the inverse projecting of a second pixel of the second image refer to a common generated point, and responsively attribute a color value to the common generated point by selecting a color value associated with one of the first pixel and the second pixel, wherein the associated pixel has a lowest local depth variance of the first pixel and the second pixel, wherein the local depth variance of a given pixel is computed for a region of n×n pixels around the given pixel, with n being an integer greater than or equal to three.
8. The device according to claim 7 , wherein the inverse projection of a pixel of an image is a three-dimension volume determined according to a resolution of the image in 3D space and to an imprecision value on the depth attribute of the pixel.
9. The device according to claim 8 , wherein determining that the first inverse projection of the first pixel and the second inverse projection of the second pixel refer to a common generated point comprises determining that a percentage of overlap of a volume of the first inverse projection and a volume of the second inverse projection is bigger than a threshold.
10. The device according to claim 8 , wherein determining that the first inverse projection of the first pixel and the second inverse projection of the second pixel refer to a common generated point comprises determining that a center of a volume of the first inverse projection is within a volume of the second inverse projection or that a center of the volume of the second inverse projection is within the volume of the first inverse projection.
11. The device according to claim 7 , wherein depth attributes and color attributes of at least one of the first and the second image are stored in a pair of one depth image and one texture image.
12. The device according to claim 7 , comprising decoding the first and second images from a received bitstream.
13. A non-transitory processor-readable storage medium having stored thereon instructions that are operative, when executed by at least one processor, to cause the at least one processor to:
obtaining a first image and a second image, the first and second images encoding a projection of an original three-dimensional colored point cloud, pixels of the first and second images comprising a depth attribute and a color attribute;
inverse projecting pixels of the first image and inverse projecting pixels of the second image; and
determining that the inverse projecting of a first pixel of the first image and the inverse projecting of a second pixel of the second image refer to a common generated point, and responsively attributing a color value to the common generated point, by selecting a color value associated with one of the first pixel and the second pixel, wherein the associated pixel has a lowest local depth variance of the first pixel and the second pixel, wherein the local depth variance of a given pixel is computed for a region of n×n pixels around the given pixel, with n being an integer greater than or equal to three.