Occlusion detection
An apparatus and method provide for occlusion detection. An example apparatus includes means for: receiving first data comprising a reference viewpoint of a real-world space; receiving second data comprising a position of a target object in the real-world space; transforming, based on the first and second data, at least one of the reference viewpoint and the target object position into a common virtual reference space; generating one or more rays which extend between the reference viewpoint and a respective spatial point associated with the position of the target object in the common virtual reference space; and determining the digital model representing the real-world space and including one or more real-world features, an occlusion status between the reference viewpoint and the target object. The occlusion status is based on an intersection of the one or more real-world features and the one or more generated rays.
1. An apparatus, comprising:
at least one processor; and
at least one memory storing instructions which, when executed by the at least one processor, cause the apparatus to:
receive first data comprising a reference viewpoint of a real-world space;
receive second data comprising a position of a target object in the real-world space, wherein the second data further comprises a three-dimensional volume associated with the target object;
transform, based on the first and second data, at least one of the reference viewpoint and the target object position into a common virtual reference space;
spatially sample a plurality of points of the three-dimensional volume to determine one or more spatial points associated with the position of the target object;
generate one or more rays which extend between the reference viewpoint and a respective one of the one or more spatial points associated with the position of the target object in the common virtual reference space, wherein generating one or more rays comprises generating a plurality of rays extending between the reference viewpoint and each of the plurality of sampled points of the volume; and
determine, using the generated one or more rays and a digital model, the digital model representing the real-world space and including one or more real-world features thereof, an occlusion status between the reference viewpoint and the target object,
wherein the occlusion status is based on a number of the plurality of rays intersected by the one or more real-world features.
2. The apparatus of claim 1 , wherein the common virtual reference space corresponds to, or maps to, the reference space of the digital model.
3. The apparatus of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the apparatus to transform the reference space of the digital model into the common virtual reference space.
4. The apparatus of claim 1 , wherein the one or more real-world features of the digital model comprise static features of the real-world space.
5. The apparatus of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the apparatus to receive a weight map comprising a plurality of weights, each of the plurality of weights associated with a respective portion of the three-dimensional volume,
wherein, for each respective portion of the three-dimensional volume, the sampling comprises sampling a respective portion of the plurality of points.
6. The apparatus of claim 5 , wherein the sampling comprises sampling the respective portion of the plurality of points using a sampling rate which is reflective of the weight associated with the respective portion of the three-dimensional volume.
7. The apparatus of claim 5 , wherein the instructions, when executed by the at least one processor, further cause the apparatus to generate a two-dimensional projection of the weight map based on a two-dimensional projection of the three-dimensional volume from the reference viewpoint,
wherein the sampling comprises sampling the respective portion of the plurality of points using a sampling rate determined based on the two-dimensional projection of the weight map.
8. The apparatus of claim 5 , wherein the second data is received from one or more sensors, wherein one or more of the plurality of weights of the weight map are indicative of a margin of error of the one or more sensors.
9. The apparatus of claim 1 , wherein:
the sampling comprises spatially sampling the plurality of points of the three-dimensional volume at a uniform sampling rate; or
the sampling comprises spatially sampling the plurality of points of the three-dimensional volume at a sampling rate determined based on a two-dimensional projection of the three-dimensional volume from the reference viewpoint; or
the sampling comprises randomly sampling the plurality of points, optionally configured to randomly sample the plurality of points using a Monte Carlo method.
10. The apparatus of claim 1 , wherein the instructions, when executed by the at least one processor, further cause the apparatus to generate an occlusion vector by transforming each of the one or more generated rays into a reference space of the reference viewpoint,
the occlusion vector comprising, for each of the one or more generated rays and the respective spatial point associated with the position of the target object, a transformation of the respective spatial point into the reference space of the reference viewpoint and an indication of whether the ray is occluded.
11. The apparatus of claim 10 , wherein the instructions, when executed by the at least one processor, further cause the apparatus to output the occlusion vector.
12. A method comprising:
receiving first data comprising a reference viewpoint of a real-world space;
receiving second data comprising a position of a target object in the real-world space, wherein the second data further comprises a three-dimensional volume associated with the target object;
transforming, based on the first and second data, at least one of the reference viewpoint and the target object position into a common virtual reference space;
spatially sampling a plurality of points of the three-dimensional volume to determine one or more spatial points associated with the position of the target object;
generating one or more rays which extend between the reference viewpoint and a respective one of the one or more spatial points associated with the position of the target object in the common virtual reference space, wherein generating one or more rays comprises generating a plurality of rays extending between the reference viewpoint and each of the plurality of sampled points of the volume; and
determining, using the generated one or more rays and a digital model, the digital model representing the real-world space and including one or more real-world features thereof, an occlusion status between the reference viewpoint and the target object,
wherein the occlusion status is based on a number of the plurality of rays intersected by the one or more real-world features.
13. The method of claim 12 , further comprising generating an occlusion vector by transforming each of the one or more generated rays into a reference space of the reference viewpoint, the occlusion vector comprising, for each of the one or more generated rays and the respective spatial point associated with the position of the target object, a transformation of the respective spatial point into the reference space of the reference viewpoint and an indication of whether the ray is occluded.
14. The method of claim 12 , wherein the common virtual reference space corresponds to, or maps to, the reference space of the digital model.
15. The method of claim 12 , further comprising transforming the reference space of the digital model into the common virtual reference space.
16. The method of claim 12 , wherein the one or more real-world features of the digital model comprise static features of the real-world space.
17. The method of claim 12 , further comprising receiving a weight map comprising a plurality of weights, each of the plurality of weights associated with a respective portion of the three-dimensional volume,
wherein, for each respective portion of the three-dimensional volume, the sampling comprises sampling a respective portion of the plurality of points.
18. A non-transitory computer readable medium comprising program instructions that, when executed by an apparatus, cause the apparatus to perform at least the following:
receiving first data comprising a reference viewpoint of a real-world space;
receiving second data comprising a position of a target object in the real-world space, wherein the second data further comprises a three-dimensional volume associated with the target object;
transforming, based on the first and second data, at least one of the reference viewpoint and the target object position into a common virtual reference space;
spatially sampling a plurality of points of the three-dimensional volume to determine one or more spatial points associated with the position of the target object;
generating one or more rays which extend between the reference viewpoint and a respective one of the one or more spatial points associated with the position of the target object in the common virtual reference space, wherein generating one or more rays comprises generating a plurality of rays extending between the reference viewpoint and each of the plurality of sampled points of the volume; and
determining, using the generated one or more rays and a digital model, the digital model representing the real-world space and including one or more real-world features thereof, an occlusion status between the reference viewpoint and the target object,
wherein the occlusion status is based on a number of the plurality of rays intersected by the one or more real-world features.