Concave surface modeling in image-based visual hull
View Patent ↗Apparatus and methods disclosed herein provide for a set of reference images obtained from a camera and a reference image obtained from a viewpoint to capture an entire concave region of an object; a silhouette processing module for obtaining a silhouette image of the concave region of the object; and a virtual-image synthesis module connected to the silhouette processing module for synthesizing a virtual inside-out image of the concave region from the computed silhouette images and for generating a visual hull of the object having the concave region.
1. A method, comprising:
obtaining, at a computing device, a plurality of reference images of an object having a concave region; and
generating, at the computing device, a virtual inside-out image for the concave region based on the plurality of reference images by:
obtaining a visual hull for an outside boundary of the object based on the plurality of reference images;
determining three-dimensional coordinates of a boundary of the concave region on the visual hull;
selecting a position inside the concave region for calculating the virtual inside-out image for the concave region;
determining image parameters for the position;
capturing a reference concave region image comprising the boundary of the concave region; and
generating the virtual inside-out image for the concave region based on the three-dimensional coordinates, the image parameters, and the reference concave region image by performing three-dimensional to two-dimensional perspective projection.
2. The method of claim 1 , wherein determining the three-dimensional coordinates of the boundary of the concave region comprises:
determining the boundary of the concave region in a two-dimensional reference picture by image segmentation; and
determining the three-dimensional coordinates of the boundary of the concave region on the visual hull using the boundary of the concave region in the two-dimensional reference picture.
3. The method of claim 1 , further comprising obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to the virtual inside-out image.
4. The method of claim 1 , wherein generating the virtual inside-out image comprises generating a plurality of virtual inside-out images, and wherein the method further comprises obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to each of the plurality of virtual inside-out images.
5. A non-transitory computer-readable storage medium storing a computer executable program code that, when executed by a computer, performs a method comprising:
obtaining a plurality of reference images of an object having a concave region; and
generating a virtual inside-out image for the concave region based on the plurality of reference images by effectuating operations comprising:
obtaining a visual hull for an outside boundary of the object based on the plurality of reference images;
determining three-dimensional coordinates of a boundary of the concave region on the visual hull;
selecting a position inside the concave region for calculating the virtual inside-out image for the concave region;
determining image parameters for the position;
capturing a reference concave region image comprising the boundary of the concave region; and
generating the virtual inside-out image for the concave region based on the three-dimensional coordinates, the image parameters, and the reference concave region image by performing three-dimensional to two-dimensional perspective protection.
6. The non-transitory computer-readable storage medium of claim 5 , wherein determining the three-dimensional coordinates of the boundary of the concave region comprises:
determining the boundary of the concave region in a two-dimensional reference picture by image segmentation; and
determining the three-dimensional coordinates of the boundary of the concave region on the visual hull using the boundary of the concave region in the two-dimensional reference picture.
7. The non-transitory computer-readable storage medium of claim 5 , wherein the method further comprises obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to the virtual inside-out image.
8. The non-transitory computer-readable storage medium of claim 5 , wherein generating the virtual inside-out image comprises generating a plurality of virtual inside-out images, and wherein the method further comprises obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to each of the plurality of virtual inside-out images.
9. An apparatus, comprising:
a storage medium configured to store a plurality of reference images of an object;
a processor coupled to the storage medium, wherein the processor is configured to generate a virtual inside-out image for a concave region based on the plurality of reference images by effectuating operations comprising:
obtaining a visual hull for an outside boundary of the object based on the plurality of reference images;
determining three-dimensional coordinates of a boundary of the concave region on the visual hull;
selecting a position inside the concave region for calculating the virtual inside-out image for the concave region;
determining image parameters for the position;
capturing a reference concave region image comprising the boundary of the concave region; and
generating the virtual inside-out image for the concave region based on the three-dimensional coordinates, the image parameters, and the reference concave region image by performing three-dimensional to two-dimensional perspective protection.
10. The apparatus of claim 9 , wherein determining the image parameters for the position comprises determining virtual camera parameters comprising focal distance.
11. The apparatus of claim 10 , wherein the position is located at the bottom of the concave region of the object.
12. The apparatus of claim 9 , wherein operations further comprise obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to the virtual inside-out image.
13. The apparatus of claim 9 , wherein the operations further comprise:
generating a plurality of virtual inside-out images, and
obtaining a revised visual hull by performing a subtraction Boolean operation for segments mapped from each of the plurality of reference images to each of the plurality of virtual inside-out images.
14. The method of claim 1 , wherein determining the image parameters for the position comprises determining virtual camera parameters comprising focal distance.
15. The non-transitory computer-readable storage medium of claim 5 , wherein determining the image parameters for the position comprises determining virtual camera parameters comprising focal distance.
16. The apparatus of claim 9 , wherein determining the three-dimensional coordinates of the boundary of the concave region comprises:
determining the boundary of the concave region in a two-dimensional reference picture by image segmentation; and
determining the three-dimensional coordinates of the boundary of the concave region on the visual hull using the boundary of the concave region in the two-dimensional reference picture.