Systems and methods for machine perception
A system to determine a position of one or more objects includes a transmitter to emit a beam of photons to sequentially illuminate regions of one or more objects; multiple cameras that are spaced-apart with each camera having an array of pixels to detect photons; and one or more processor devices that execute stored instructions to perform actions of a method, including: directing the transmitter to sequentially illuminate regions of one or more objects with the beam of photons; for each of the regions, receiving, from the cameras, an array position of each pixel that detected photons of the beam reflected or scattered by the region of the one or more objects; and, for each of the regions detected by the cameras, determining a position of the regions using the received array positions of the pixels that detected the photons of the beam reflected or scattered by that region.
1. A system to determine a position of one or more objects, comprising:
one or more transmitters arranged to emit four or more beams of photons to scan a plurality of regions that include the one or more objects;
three or more cameras arranged to detect one or more reflected beams of photons from each voxel, wherein each voxel represents a sampled surface element of a three-dimensional shaped surface of the plurality of regions; and
one or more processor devices that execute instructions to enable performance actions, including:
employing a position of each pixel in the three or more cameras to detect the beam of photons reflected by each voxel for one or more of the plurality of objects; and
determining each three-dimensional position of the one or more objects using the position of each pixel in the three or more cameras.
2. The system of claim 1 , further comprising:
employing each reflection of the four or more beams of photons to determine a position for each of the three or more cameras.
3. The system of claim 1 , further comprising:
employing linear interpolation of each detected reflection of the beam of photons to improve spatial accuracy and temporal resolution of each pixel for the three or more cameras.
4. The system of claim 1 , further comprising:
employing overlapping field of views (FOVs) of one or more portions of the reflected beams of photons to determine a volumetric shape of each object positioned within the overlapping FOVs.
5. The system of claim 1 , further comprising:
a laser brush that is arranged for manually illuminating an object with a scanning laser beam that creates brush strokes of beams of photons on one or more surfaces of the object.
6. The system of claim 1 , further comprising:
automatically selecting three focal perspectives for reflected beams of photons based on an amount of concurrency with each associated field of view.
7. The system of claim 1 , further comprising:
employing one or more appendages on a head of user to provide spacing of a position for each of the three or more cameras away from each of the other cameras.
8. The system of claim 1 , further comprising:
integrating the actions of claim 1 as a subsystem for one or more of an augmented reality system or a virtual reality system.
9. A method to determine a position of one or more objects, comprising:
employing one or more transmitters arranged to emit four or more beams of photons to scan a plurality of regions that include the one or more objects;
employing three or more cameras arranged to detect one or more reflected beams of photons from each voxel, wherein each voxel represents a sampled surface element of a three-dimensional shaped surface of the plurality of regions; and
employing one or more processor devices to execute instructions that enable performance actions, including:
employing a position of each pixel in the three or more cameras to detect the beam of photons reflected by each voxel for one or more of the plurality of objects; and
determining each three-dimensional position of the one or more objects using the position of each pixel in the three or more cameras.
10. The method of claim 9 , further comprising:
employing each reflection of the four or more beams of photons to determine a position for each of the three or more cameras.
11. The method of claim 9 , further comprising:
employing linear interpolation of each detected reflection of the beam of photons to improve spatial accuracy and temporal resolution of each pixel for the three or more cameras.
12. The method of claim 9 , further comprising:
employing overlapping field of views (FOVs) of one or more portions of the reflected beams of photons to determine a volumetric shape of each object positioned within the overlapping FOVs.
13. The method of claim 9 , further comprising:
a laser brush that is arranged for manually illuminating an object with a scanning laser beam that creates brush strokes of beams of photons on one or more surfaces of the object.
14. The method of claim 9 , further comprising:
automatically selecting three focal perspectives for reflected beams of photons based on an amount of concurrency with each associated field of view.
15. The method of claim 9 , further comprising:
employing one or more appendages on a head of user to provide spacing of a position for each of the three or more cameras away from each of the other cameras.
16. The method of claim 9 , further comprising:
integrating the actions of claim 1 as a subsystem for one or more of an augmented reality system or a virtual reality system.
17. A computer readable non-transitory storage media that includes instructions for determining a position of one or more objects, wherein execution of the instructions by one or more processors enables actions, comprising:
employing one or more transmitters arranged to emit four or more beams of photons to scan a plurality of regions that include the one or more objects;
employing three or more cameras arranged to detect one or more reflected beams of photons from each voxel, wherein each voxel represents a sampled surface element of a three-dimensional shaped surface of the plurality of regions; and
employing one or more processor devices to execute instructions that enable performance actions, including:
employing a position of each pixel in the three or more cameras to detect the beam of photons reflected by each voxel for one or more of the plurality of objects; and
determining each three-dimensional position of the one or more objects using the position of each pixel in the three or more cameras.
18. The computer readable non-transitory storage media of claim 17 , further comprising:
employing each reflection of the four or more beams of photons to determine a position for each of the three or more cameras.
19. The computer readable non-transitory storage medium method of claim 17 , further comprising:
employing linear interpolation of each detected reflection of the beam of photons to improve spatial accuracy and temporal resolution of each pixel for the three or more cameras.
20. The computer readable non-transitory storage medium of claim 17 , further comprising:
employing overlapping field of views (FOVs) of one or more portions of the reflected beams of photons to determine a volumetric shape of each object positioned within the overlapping FOVs.