Image processing apparatus and image processing method for 3D data compression
The present disclosure relates to an image processing apparatus and an image processing method capable of suppressing a reduction in encoding efficiency. A bit stream is generated which contains: information indicating a correspondence relation between at least one of a geometry image obtained by projecting position information regarding 3D data representing a three-dimensional structure onto a two-dimensional plane or a texture image obtained by projecting attribute information regarding the 3D data onto a two-dimensional plane, and an occupancy map that is map information indicating whether or not data is present at each position; and encoded data regarding the geometry image, encoded data regarding the texture image, and encoded data regarding the occupancy map. The present disclosure is applicable to, for example, an information processing apparatus, an image processing apparatus, an electronic apparatus, an information processing method, a program, and the like.
1. An image encoding apparatus, comprising:
circuitry configured to:
obtain a geometry image, a texture image, and an occupancy map by projecting patch images of 3D data representing a three-dimensional structure onto respective two-dimensional planes, wherein
the occupancy map is map information indicating that data is present at each position, and
at least one of the geometry image or the texture image has a resolution different from a resolution of the occupancy map;
encode the geometry image, the texture image, and the occupancy map; and
generate a bit stream comprising the encoded geometry image, the encoded texture image, the encoded occupancy map, a first variable indicating a first relation between the resolution of the occupancy map and the resolution of the geometry image, and a second variable indicating a second relation between the resolution of the occupancy map and the resolution of the texture image, wherein
the first variable and the second variable indicate correspondence information between a range of each of the patch images of the occupancy map and a range of each of the patch images of at least one of the geometry image or the texture image.
2. The image encoding apparatus according to claim 1 , wherein each of the first variable and the second variable is defined in a horizontal direction and a vertical direction.
3. An image encoding method, comprising:
obtaining a geometry image, a texture image, and an occupancy map by projecting patch images of 3D data representing a three-dimensional structure onto respective two-dimensional planes, wherein
the occupancy map is map information indicating that data is present at each position, and
at least one of the geometry image or the texture image has a resolution different from a resolution of the occupancy map;
encoding the geometry image, the texture image, and the occupancy map; and
generating a bit stream comprising the encoded geometry image, the encoded texture image, the encoded occupancy map,
a first variable indicating a first relation between the resolution of the occupancy map and the resolution of the geometry image, and
a second variable indicating a second relation between the resolution of the occupancy map and the resolution of the texture image, wherein
the first variable and the second variable indicate correspondence information between a range of each of the patch images of the occupancy map and a range of each of the patch images of at least one of the geometry image or the texture image.
4. An image decoding apparatus, comprising:
circuitry configured to:
obtain, from a bit stream, an encoded geometry image, an encoded texture image, an encoded occupancy map, a first variable, and a second variable, wherein
the encoded occupancy map is map information indicating whether data is present at each position,
the first variable indicates a first relation between a resolution of the encoded occupancy map and a resolution of the encoded geometry image, and
the second variable indicates a second relation between the resolution of the encoded occupancy map and a resolution of the encoded texture image;
decode the encoded geometry image, the encoded texture image, and the encoded occupancy map;
change, based on the first variable and the second variable, the resolution of the decoded geometry image and the resolution of the decoded texture image to be identical to the resolution of the decoded occupancy map;
unpack, the decoded geometry image and the decoded texture image having the resolution identical to that of the decoded occupancy map; and
reconstruct 3D data based on the unpacked geometry image, the unpacked texture image, and the decoded occupancy map, wherein the 3D data represents a three-dimensional structure, wherein
the first variable and the second variable indicate correspondence information between a range of each of patch images of the encoded occupancy map and a range of each of patch images of at least one of the encoded geometry image or the encoded texture image.
5. The image decoding apparatus according to claim 4 , wherein each of the first variable and the second variable is defined in a horizontal direction and a vertical direction.
6. The image decoding apparatus according to claim 4 , wherein the 3D data includes a point cloud.
7. An image decoding method, comprising:
obtaining, from a bit stream, an encoded geometry image, an encoded texture image, an encoded occupancy map, a first variable, and a second variable, wherein
the encoded occupancy map is map information indicating whether data is present at each position,
the first variable indicates a first relation between a resolution of the encoded occupancy map and a resolution of the encoded geometry image, and
the second variable indicates a second relation between the resolution of the encoded occupancy map and a resolution of the encoded texture image;
decoding the encoded geometry image, the encoded texture image, and the encoded occupancy map;
changing, based on the first variable and the second variable, the resolution of the decoded geometry image and the resolution of the decoded texture image to be identical to the resolution of the decoded occupancy map;
unpacking, the decoded geometry image and the decoded texture image having the resolution identical to that of the decoded occupancy map; and
reconstructing 3D data based on the unpacked geometry image, the unpacked texture image, and the decoded occupancy map, wherein the 3D data represents a three-dimensional structure, wherein
the first variable and the second variable indicate correspondence information between a range of each of patch images of the encoded occupancy map and a range of each of patch images of at least one of the encoded geometry image or the encoded texture image.