Virtual-environment-based object construction method and apparatus, computer device, and computer-readable storage medium
A virtual-environment-based object construction method and apparatus, and a computer-readable storage medium are disclosed in this disclosure, relating to the field of virtual environment. Performed by a terminal comprising a camera, the method including: displaying an environment interface; receiving a capture operation being used for acquiring three-dimensional (3D) information of a to-be-acquired object; receiving a position input operation for determining a display position of the target object in the virtual environment; and displaying the target object at the display position in the virtual environment according to the capture operation and the position input operation.
1. A virtual-environment-based object construction method, performed by a terminal, the method comprising:
displaying an environment interface, the environment interface comprising an image corresponding to a virtual environment;
obtaining a feature point cloud from a capture operation performed by a camera, the feature point cloud determining a style of a target object to be displayed;
receiving a position input operation indicating a display position of the target object in the virtual environment; and
displaying the target object at the display position in the virtual environment by filling a region corresponding to the feature point cloud with voxel blocks, wherein filling the region corresponding to the feature point cloud with the voxel blocks comprises:
performing three-dimensional (3D) slicing on a bounding box corresponding to the feature point cloud to obtain voxel regions; and
filling the voxel regions with the voxel blocks according to an inclusion relationship between the voxel regions and pixels in the feature point cloud by filling the voxel regions with the voxel blocks in response to a quantity of the pixels in the voxel regions being greater than a preset quantity.
2. The method according to claim 1 , further comprising:
receiving the capture operation to capture n frames of images around a to-be-acquired object, and capturing the n frames of images, n being a positive integer; and
generating the feature point cloud according to depth information corresponding to each of the n frames of images, or performing 3D reconstruction on the to-be-acquired object based on the n frames of images to obtain the feature point cloud.
3. The method according to claim 2 , wherein generating the feature point cloud according to the depth information corresponding to the each of the n frames of images comprises:
determining a relative position of the camera during capturing of the each of the n frames of images, the relative position being determined according to a relationship with a position in which the camera captures a key-frame image;
determining positions of pixels in a 3D space according to positions of the pixels in the each of the n frames of images, the depth information corresponding to the pixels, and the relative position of the camera; and
obtaining the feature point cloud according to the positions of the pixels in the 3D space.
4. The method according to claim 2 , wherein receiving the capture operation for capturing the n frames of images around the to-be-acquired object comprises one of:
receiving a video capture operation around the to-be-acquired object, a video stream captured by the video capture operation comprising the n frames of images; or
receiving a fixed-point capture operation around the to-be-acquired object, the fixed-point capture operation being used for capturing the n frames of images at designated positions around the to-be-acquired object.
5. The method according to claim 1 , further comprising:
receiving a 3D slicing operation to obtain a slicing mode corresponding to each dimension of 3D dimensions, the 3D slicing operation being used to perform the 3D slicing on the bounding box corresponding to the feature point cloud according to the slicing mode.
6. The method according to claim 5 , wherein receiving the 3D slicing operation comprises:
receiving slice quantities from a slice quantity input operation, the slice quantities corresponding to three dimensions of the feature point cloud; and performing the 3D slicing on the bounding box based on the slice quantities; or
receiving a sliding slicing operation indicating the slice quantities, and performing the 3D slicing on the bounding box according to the slice quantities, the slice quantities corresponding to three dimensions being used to determine a degree of refinement of the target object generated by the feature point cloud.
7. The method according to claim 1 , wherein filling the voxel regions with the voxel blocks comprises:
determining a weighted mean color of the pixels in the voxel regions to obtain a target color, and filling the voxel regions with the voxel blocks having a color closest to the target color.
8. The method according to claim 1 , wherein filling the voxel regions with the voxel blocks comprises:
determining a color with a highest proportion in distribution as a target color according to color distribution of the pixels in the voxel regions, and filling the voxel regions with the voxel blocks having a color closest to the target color.
9. A device for virtual-environment-based object construction, the device comprising:
a memory storing a plurality of computer instructions; and
a processor configured to execute the plurality of computer instructions, wherein upon execution of the plurality of computer instructions, the processor is configured to cause the device to:
display an environment interface, the environment interface comprising an image corresponding to a virtual environment;
obtain a feature point cloud from a capture operation performed by a camera, the feature point cloud determining a style of a target object to be displayed;
receive a position input operation indicating a display position of the target object in the virtual environment; and
display the target object at the display position in the virtual environment by filling a region corresponding to the feature point cloud with voxel blocks, wherein filling the region corresponding to the feature point cloud with the voxel blocks comprises:
performing three-dimensional (3D) slicing on a bounding box corresponding to the feature point cloud to obtain voxel regions; and
filling the voxel regions with the voxel blocks according to an inclusion relationship between the voxel regions and pixels in the feature point cloud by filling the voxel regions with the voxel blocks in response to a quantity of the pixels in the voxel regions being greater than a preset quantity.
10. The device according to claim 9 , wherein, when the processor, upon execution of the plurality of computer instructions, is further configured to cause the device to:
receive the capture operation to capture n frames of images around a to-be-acquired object, and capture the n frames of images, n being a positive integer; and
generate the feature point cloud according to depth information corresponding to each of the n frames of images, or perform 3D reconstruction on the to-be-acquired object based on the n frames of images to obtain the feature point cloud.
11. The device according to claim 10 , wherein in order to cause the device to generate the feature point cloud according to the depth information corresponding to the each of the n frames of images, the processor is configured to cause the device to:
determine a relative position of the camera during capturing of the each of the n frames of images, the relative position being determined according to a relationship with a position in which the camera captures a key-frame image;
determine positions of pixels in a 3D space according to positions of the pixels in the each of the n frames of images, the depth information corresponding to the pixels, and the relative position of the camera; and
obtain the feature point cloud according to the positions of the pixels in the 3D space.
12. The device according to claim 10 , wherein in order to cause the device to receive the capture operation to capture the n frames of images around the to-be-acquired object, the processor is configured to cause the device to:
receive a video capture operation around the to-be-acquired object, a video stream captured by the video capture operation comprising the n frames of images; or
receive a fixed-point capture operation around the to-be-acquired object, the fixed-point capture operation being used for capturing the n frames of images at designated positions around the to-be-acquired object.
13. The device according to claim 9 , wherein upon execution of the plurality of computer instructions, the processor is further configured to cause the device to:
receive a 3D slicing operation to obtain a slicing mode corresponding to each dimension of 3D dimensions, the 3D slicing operation being used to perform the 3D slicing on the bounding box corresponding to the feature point cloud according to the slicing mode.
14. The device according to claim 13 , wherein in order to receive the 3D slicing operation, the processor is configured to cause the device to:
receive slice quantities from a slice quantity input operation, the slice quantities corresponding to three dimensions of the feature point cloud; and perform the 3D slicing on the bounding box based on the slice quantities; or
receive a sliding slicing operation indicating the slice quantities, and perform the 3D slicing on the bounding box according to the slice quantities, the slice quantities corresponding to three dimensions being used to determine a degree of refinement of the target object generated by the feature point cloud.
15. The device according to claim 9 , wherein in order to cause the device to fill the voxel regions with the voxel blocks, the processor is configured to cause the device to:
determine a weighted mean color of the pixels in the voxel regions to obtain a target color, and fill the voxel regions with the voxel blocks having a color closest to the target color.
16. The device according to claim 9 , wherein in order to cause the device to fill the voxel regions with the voxel blocks, the processor is configured to cause the device to:
determine a color with a highest proportion in distribution as a target color according to color distribution of the pixels in the voxel regions, and fill the voxel regions with the voxel blocks having a color closest to the target color.
17. A non-transitory storage medium for storing computer readable instructions, the computer readable instructions, when executed by a processor in a device comprising a camera, causing the processor to:
display an environment interface, the environment interface comprising an image corresponding to a virtual environment;
obtain a feature point cloud from a capture operation performed by a camera, the feature point cloud determining a style of a target object to be displayed;
receive a position input operation indicating a display position of the target object in the virtual environment; and
display the target object at the display position in the virtual environment by filling a region corresponding to the feature point cloud with voxel blocks, wherein filling the region corresponding to the feature point cloud with the voxel blocks comprises:
performing three-dimensional (3D) slicing on a bounding box corresponding to the feature point cloud to obtain voxel regions; and
filling the voxel regions with the voxel blocks according to an inclusion relationship between the voxel regions and pixels in the feature point cloud by filling the voxel regions with the voxel blocks in response to a quantity of the pixels in the voxel regions being greater than a preset quantity.
18. The non-transitory storage medium according to claim 17 , wherein, when the computer readable instructions further cause the processor to:
receive the capture operation for capturing n frames of images around a to-be-acquired object, and capturing the n frames of images, n being a positive integer; and
generate the feature point cloud according to depth information corresponding to each of the n frames of images, or perform 3D reconstruction on the to-be-acquired object based on the n frames of images to obtain the feature point cloud.
19. The non-transitory storage medium according to claim 18 , wherein in order to cause the processor to generate the feature point cloud according to the depth information corresponding to the each of the n frames of images, the computer readable instructions cause the processor to:
determine a relative position of the camera during capturing of the each of the n frames of images, the relative position being determined according to a relationship with a position in which the camera captures a key-frame image;
determine positions of pixels in a 3D space according to positions of the pixels in the each of the n frames of images, the depth information corresponding to the pixels, and the relative position of the camera; and
obtain the feature point cloud according to the positions of the pixels in the 3D space.
20. The non-transitory storage medium according to claim 18 , wherein in order to cause the processor to receive the capture operation to capture the n frames of images around the to-be-acquired object, the computer readable instructions cause the processor to:
receive a video capture operation around the to-be-acquired object, a video stream captured by the video capture operation comprising the n frames of images; or
receive a fixed-point capture operation around the to-be-acquired object, the fixed-point capture operation being used for capturing the n frames of images at designated positions around the to-be-acquired object.