Systems and methods for stabilizing videos
Images with an optical field of view are captured by an image capture device. An observed trajectory of the image capture device reflects the positions of the image capture device at different moments may be determined. A capture trajectory of the image capture device reflects virtual positions of the image capture device from which video content may be generated. The capture trajectory is determined based on a subsequent portion of the observed trajectory such that a portion of the capture trajectory corresponding to a portion of the observed trajectory is determined based on a subsequent portion of the observed trajectory. Orientations of punch-outs for the images are determined based on the capture trajectory. Video content is generated based on visual content of the images within the punch-outs.
1. A system for stabilizing videos, the system comprising:
one or more physical processors configured by machine-readable instructions to:
determine an observed trajectory of an image capture device for a capture duration during which the image capture device captures images having an optical field of view, the observed trajectory reflecting positions of the image capture device at the different moments within the capture duration, the observed trajectory including a first portion corresponding to a first moment within the capture duration and a second portion corresponding to a second moment subsequent to the first moment within the capture duration;
stabilize the observed trajectory to determine a capture trajectory of the image capture device based on a look ahead of the observed trajectory, the look ahead of the observed trajectory including use of a subsequent portion of the observed trajectory to stabilize a preceding portion of the capture trajectory;
determine orientations of capture field of view for the images with respect to the optical field of view of the images based on the capture trajectory of the image capture device, the capture field of view defining a punch-out of the images to generate video content; and
generate the video content based on the punch-out of the images within the capture field of view.
2. The system of claim 1 , wherein the determination of the capture trajectory of the image capture device based on the look ahead of the observed trajectory preserves a user's intended motion of the image capture device.
3. The system of claim 2 , wherein past positions of the image capture device supply context for the user's intended motion of the image capture device in determination of the capture trajectory that preserves the user's intended motion of the image capture device.
4. The system of claim 1 , wherein the positions of the image capture device include rotational positions of the image capture device, and at least some of capture trajectory is rotationally offset from the observed trajectory.
5. The system of claim 1 , wherein the capture trajectory is determined to include a path that minimizes a combination of a time derivative and a second time derivative of rotational positions of the image capture device while respecting a set of constraints.
6. The system of claim 5 , wherein the set of constraints includes a margin constraint, the margin constraint providing a limitation on deviation of the capture trajectory from the observed trajectory, the margin constraint determined based on a difference between the optical field of view and the capture field of view.
7. The system of claim 5 , wherein the set of constraints includes a target constraint, the target constraint requiring a target in the images to be within the punch- out of the images, the target constraint determined based on positions of the target in the images.
8. The system of claim 1 , wherein the images are stored in a buffer during the capture trajectory determination such that individual ones of the images corresponding to a portion of the capture trajectory are stored in the buffer during the determination of the portion of the capture trajectory using the lookahead of the observed trajectory.
9. The system of claim 1 , wherein a preceding moment within the capture duration corresponding to the preceding portion of the capture trajectory is adjacent to a subsequent moment within the capture duration corresponding to the subsequent portion of the observed trajectory.
10. The system of claim 1 , wherein a preceding moment within the capture duration corresponding to the preceding portion of the capture trajectory is not adjacent to a subsequent moment within the capture duration corresponding to the subsequent portion of the observed trajectory.
11. A method for stabilizing videos, the method performed by a computing system including one or more processors, the method comprising:
determining, by the one or more processors, an observed trajectory of an image capture device for a capture duration during which the image capture device captures images having an optical field of view, the observed trajectory reflecting positions of the image capture device at the different moments within the capture duration, the observed trajectory including a first portion corresponding to a first moment within the capture duration and a second portion corresponding to a second moment subsequent to the first moment within the capture duration;
stabilizing, by the one or more processors, the observed trajectory to determine a capture trajectory of the image capture device based on a look ahead of the observed trajectory, the look ahead of the observed trajectory including use of a subsequent portion of the observed trajectory to stabilize a preceding portion of the capture trajectory;
determining, by the one or more processors, orientations of capture field of view for the images with respect to the optical field of view of the images based on the capture trajectory of the image capture device, the capture field of view defining a punch-out of the images to generate video content; and
generating, by the one or more processors, the video content based on the punch-out of the images within the capture field of view.
12. The method of claim 11 , wherein determining the capture trajectory of the image capture device based on the look ahead of the observed trajectory preserves a user's intended motion of the image capture device.
13. The method of claim 12 , wherein past positions of the image capture device supply context for the user's intended motion of the image capture device in determining the capture trajectory that preserves the user's intended motion of the image capture device.
14. The method of claim 11 , wherein the positions of the image capture device include rotational positions of the image capture device, and at least some of capture trajectory is rotationally offset from the observed trajectory.
15. The method of claim 11 , wherein the capture trajectory is determined to include a path that minimizes a combination of a time derivative and a second time derivative of rotational positions of the image capture device while respecting a set of constraints.
16. The method of claim 15 , wherein the set of constraints includes a margin constraint, the margin constraint providing a limitation on deviation of the capture trajectory from the observed trajectory, the margin constraint determined based on a difference between the optical field of view and the capture field of view.
17. The method of claim 15 , wherein the set of constraints includes a target constraint, the target constraint requiring a target in the images to be within the punch-out of the images, the target constraint determined based on positions of the target in the images.
18. The method of claim 11 , wherein the images are stored in a buffer during the capture trajectory determination such that individual ones of the images corresponding to a portion of the capture trajectory are stored in the buffer during the determination of the portion of the capture trajectory using the lookahead of the observed trajectory.
19. The method of claim 11 , wherein a preceding moment within the capture duration corresponding to the preceding portion of the capture trajectory is adjacent to a subsequent moment within the capture duration corresponding to the subsequent portion of the observed trajectory.
20. The method of claim 11 , wherein a preceding moment within the capture duration corresponding to the preceding portion of the capture trajectory is not adjacent to a subsequent moment within the capture duration corresponding to the subsequent portion of the observed trajectory.