Determining a virtual representation of an environment by projecting texture patterns
Example methods and systems for determining 3D scene geometry by projecting patterns of light onto a scene are provided. In an example method, a first projector may project a first random texture pattern having a first wavelength and a second projector may project a second random texture pattern having a second wavelength. A computing device may receive sensor data that is indicative of an environment as perceived from a first viewpoint of a first optical sensor and a second viewpoint of a second optical sensor. Based on the received sensor data, the computing device may determine corresponding features between sensor data associated with the first viewpoint and sensor data associated with the second viewpoint. And based on the determined corresponding features, the computing device may determine an output including a virtual representation of the environment that includes depth measurements indicative of distances to at least one object.
1. A method comprising:
causing a projector to project a plurality of different patterns of light during a given time period;
receiving sensor data comprising a first sequence of infrared images corresponding to the given time period and captured from a first viewpoint of a first optical sensor and a second sequence of infrared images corresponding to the given time period and captured from a second viewpoint of a second optical sensor;
determining a first spatio-temporal image based on a combination of the first sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the first sequence of images;
determining a second spatio-temporal image based on a combination of the second sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the second sequence of images;
determining, by the computing device, corresponding features between the first spatio-temporal image and the second spatio-temporal image; and
based on the determined corresponding features, determining, by the computing device, an output including a virtual representation of an environment, wherein the output comprises a depth measurement indicative of a distance from the first optical sensor to at least one object in the environment.
2. The method of claim 1 , wherein the determined corresponding features comprise corresponding segments of pixels.
3. The method of claim 1 , wherein projecting the plurality of different patterns of light comprises projecting a first texture pattern during a first portion of the given time period and projecting a second texture pattern during a second portion of the given time period.
4. The method of claim 3 , wherein:
the first sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern, and
the second sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern.
5. The method of claim 1 , wherein:
the first spatio-temporal image is a data structure that stores information about changes in images of the first sequence of infrared images over time, and
the second spatio-temporal image is a data structure that stores information about changes in images of the second sequence of infrared images over time.
6. The method of claim 1 , wherein the first optical sensor and the second optical sensor are coupled to a robotic manipulator.
7. The method of claim 6 , wherein the projector is coupled to the robotic manipulator.
8. The method of claim 1 , wherein the projector is coupled to a robotic manipulator.
9. The method of claim 1 , wherein:
determining the first spatio-temporal image comprises summing the plurality of different patterns of light projected during the given time period and captured in the first sequence of infrared images; and
determining the second spatio-temporal image comprises summing the plurality of different patterns of light projected during the given time period and captured in the second sequence of infrared images.
10. A system comprising:
a projector;
a first optical sensor;
a second optical sensor; and
a computing device comprising one or more processors and a memory, wherein the one or more processors is configured to perform functions comprising:
causing the projector to project a plurality of different patterns of light during a given time period;
receiving sensor data comprising a first sequence of infrared images corresponding to the given time period and captured from a first viewpoint of the first optical sensor and a second sequence of infrared images corresponding to the given time period and captured from a second viewpoint of the second optical sensor;
determining a first spatio-temporal image based on a combination of the first sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the first sequence of images;
determining a second spatio-temporal image based on a combination of the second sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the second sequence of images;
determining corresponding features between the first spatio-temporal image and the second spatio-temporal image; and
based on the determined corresponding features, determining an output including a virtual representation of an environment, wherein the output comprises a depth measurement indicative of a distance from the first optical sensor to at least one object in the environment.
11. The system of claim 10 , wherein the determined corresponding features comprise corresponding segments of pixels.
12. The system of claim 10 , wherein projecting the plurality of different patterns of light comprises projecting a first texture pattern during a first portion of the given time period and projecting a second texture pattern during a second portion of the given time period.
13. The system of claim 12 , wherein:
the first sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern, and
the second sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern.
14. The system of claim 10 , wherein:
the first spatio-temporal image is a data structure that stores information about changes in images of the first sequence of infrared images over time, and
the second spatio-temporal image is a data structure that stores information about changes in images of the second sequence of infrared images over time.
15. The system of claim 10 , further comprising a robotic manipulator,
wherein the first optical sensor and the second optical sensor are coupled to the robotic manipulator.
16. The system of claim 10 , further comprising a robotic manipulator, wherein the projector is coupled to the robotic manipulator.
17. A non-transitory computer-readable medium having stored therein instructions, that when executed by a computing device, cause the computing device to perform functions comprising:
causing the projector to project a plurality of different patterns of light during a given time period;
receiving sensor data comprising a first sequence of infrared images corresponding to the given time period and captured from a first viewpoint of the first optical sensor and a second sequence of infrared images corresponding to the given time period and captured from a second viewpoint of the second optical sensor;
determining a first spatio-temporal image based on a combination of the first sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the first sequence of images;
determining a second spatio-temporal image based on a combination of the second sequence of infrared images, wherein the combination comprises a plurality of footprints, wherein each footprint of the plurality describes a corresponding segment of pixels by combining information over the given time period for the corresponding segment of pixels from each image of the second sequence of images;
determining corresponding features between the first spatio-temporal image and the second spatio-temporal image; and
based on the determined corresponding features, determining an output including a virtual representation of an environment, wherein the output comprises a depth measurement indicative of a distance from the first optical sensor to at least one object in the environment.
18. The non-transitory computer-readable medium of claim 17 , wherein the determined corresponding features comprise corresponding segments of pixels.
19. The non-transitory computer-readable medium of claim 17 , wherein projecting the plurality of different patterns of light comprises projecting a first texture pattern during a first portion of the given time period and projecting a second texture pattern during a second portion of the given time period.
20. The non-transitory computer-readable medium of claim 19 , wherein: the first sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern, and
the second sequence of infrared images comprises at least one image of the first texture pattern and at least one image of the second texture pattern.