Camera view synthesis on head-mounted display for virtual reality and augmented reality
Various examples pertaining to camera view synthesis on head-mounted display (HMD) for virtual reality (VR) and augmented reality (AR) are described. A method involves receiving, from a plurality of tracking cameras disposed around a HMD, image data of a scene which is on a first side of the HMD. The method also involves performing, using the image data and depth information pertaining to the scene, view synthesis to generate a see-through effect of viewing the scene from a viewing position on a second side of the HMD opposite the first side thereof.
1. A method, comprising:
receiving, from a plurality of tracking cameras of a head-mounted display (HMD), image data of a scene from a plurality of camera positions, respectively, and
performing, using the image data and a depth information pertaining to the scene, view synthesis to generate a see-through effect of viewing the scene from at least one viewing position different from the plurality of camera positions, the at least one viewing position being displaced along a face of the HMD from the plurality of camera positions,
wherein the at least one viewing position is more central to the HMD than the plurality of camera positions.
2. The method of claim 1 , wherein the performing of the view synthesis to generate the see-through effect of viewing the scene further comprises:
selecting one of a plurality of predefined viewing positions each of which is a fixed position; and
synthesizing viewing the scene from the selected one of the plurality of predefined viewing positions.
3. The method of claim 1 , wherein the performing of the view synthesis to generate the see-through effect of viewing the scene further comprises generating an image of the scene with an effect of viewing the scene via one or more see-through cameras located corresponding to eyes of a user wearing the HMD.
4. The method of claim 1 , wherein the performing of the view synthesis to generate the see-through effect of viewing the scene further comprises synthesizing viewing of the scene via one or more see-through cameras by performing depth image-based rendering (DIBR) with hole filling.
5. The method of claim 1 , further comprising:
generating depth signals using the image data from the plurality of tracking cameras; and
determining the depth information based on the depth signals.
6. The method of claim 1 , wherein the plurality of tracking cameras comprises at least a first tracking camera with a first frame rate and a second tracking camera with a second frame rate different from the first frame rate.
7. The method of claim 6 , wherein performing, using the image data and depth information pertaining to the scene, view synthesis to generate the see-through effect of viewing the scene from the at least one viewing position generates images at the first frame rate, the second frame rate or a third frame rate.
8. The method of claim 1 , wherein the plurality of tracking cameras comprises at least a first tracking camera with a first resolution and a second tracking camera with a second resolution different from the first resolution.
9. The method of claim 8 , wherein performing, using the image data and depth information pertaining to the scene, view synthesis to generate the see-through effect of viewing the scene from the at least one viewing position generates images at the first resolution, the second resolution or a third resolution.
10. An apparatus, comprising:
a head-mounted display (HMD) comprising:
a head-mount piece wearable by a user on a head of the user;
a plurality of tracking cameras disposed around the head-mount piece; and
a processor configured to, during operation, perform operations comprising:
receiving, from the plurality of tracking cameras, image data of a scene from a plurality of camera positions, respectively; and
performing, using the image data and a depth information pertaining to the scene, view synthesis to generate a see-through effect of viewing the scene from at least one viewing position different from the plurality of camera positions, the at least one viewing position being displaced along a face of the HMD from the plurality of camera positions,
wherein the at least one viewing position is more central to the HMD than the plurality of camera positions.
11. The apparatus of claim 10 , wherein, in performing the view synthesis to generate the see-through effect of viewing the scene, the processor further performs operations comprising:
selecting one of a plurality of predefined viewing positions each of which is a fixed position; and
synthesizing viewing the scene from the selected one of the plurality of predefined viewing positions.
12. The apparatus of claim 10 , wherein, in performing the view synthesis to generate the see-through effect of viewing the scene, the processor further generates an image of the scene with an effect of viewing the scene via one or more see-through cameras located corresponding to eyes of a user wearing the HMD.
13. The apparatus of claim 10 , wherein, in performing the view synthesis to generate the see-through effect of viewing the scene, the processor further synthesizes viewing of the scene via one or more see-through cameras by performing depth image-based rendering (DIBR) with hole filling.
14. The apparatus of claim 10 , wherein, during operation, the processor further performs operations comprising:
generating depth signals using the image data from the plurality of tracking cameras; and
determining the depth information based on the depth signals.
15. The apparatus of claim 10 , wherein the plurality of tracking cameras comprises at least a first tracking camera with a first frame rate and a second tracking camera with a second frame rate different from the first frame rate.
16. The apparatus of claim 15 , wherein performing, using the image data and depth information pertaining to the scene, view synthesis to generate the see-through effect of viewing the scene from the at least one viewing position generates images at the first frame rate, the second frame rate or a third frame rate.
17. The apparatus of claim 10 , wherein the plurality of tracking cameras comprises at least a first tracking camera with a first resolution and a second tracking camera with a second resolution different from the first resolution.
18. The apparatus of claim 17 , wherein performing, using the image data and depth information pertaining to the scene, view synthesis to generate the see-through effect of viewing the scene from the at least one viewing position generates images at the first resolution, the second resolution or a third resolution.