Smoothness constraint for camera pose estimation
Disclosed are devices, systems and methods for incorporating a smoothness constraint for camera pose estimation. One method for robust camera pose estimation includes determining a first bounding box based on a previous frame, determining a second bounding box based on a current frame that is temporally subsequent to the previous frame, estimating the camera pose by minimizing a weighted sum of a camera pose function and a constraint function, where the camera pose function tracks a position and an orientation of the camera in time, and where the constraint function is based on coordinates of the first bounding box and coordinates of the second bounding box, and using the camera pose for navigating the vehicle. The method may further include generating an initial estimate of the camera pose is based on a Global Positioning System (GPS) sensor or an Inertial Measurement Unit (IMU).
1. A method for estimating a camera pose of a camera, implemented in a vehicle, the method comprising:
determining a first bounding box based on a previous frame;
determining a second bounding box based on a current frame, wherein the current frame is temporally subsequent to the previous frame;
estimating the camera pose by minimizing a weighted sum of a camera pose function and a constraint function, wherein the camera pose function tracks a position and an orientation of the camera in time, and wherein the constraint function is based on coordinates of the first bounding box and coordinates of the second bounding box; and
using at least the camera pose for navigating the vehicle.
2. The method of claim 1 , wherein the constraint function is further based on bounding boxes corresponding to each of a plurality of cars being tracked with a high level of confidence.
3. The method of claim 1 , wherein determining the first and second bounding boxes is based on an initial estimate of the camera pose.
4. The method of claim 3 , wherein the initial estimate of the camera pose is based on at least a Global Positioning System (GPS) sensor or an Inertial Measurement Unit (IMU).
5. The method of claim 1 , wherein a weight of the constraint function in the weighted sum is based on previous estimates of the camera pose.
6. The method of claim 1 , wherein the first and second bounding boxes correspond to a position of a rear of a car that is in front of the vehicle.
7. The method of claim 6 , wherein the first bounding box is based on predicting the location of the rear of the car using the previous frame, and wherein the second bounding box is based on detecting the location of the rear of the car using the current frame.
8. The method of claim 1 , wherein the first bounding box is further based on at least a focal length of the camera.
9. An apparatus for estimating a camera pose, comprising:
a processor; and
a memory comprising processor executable code, the processor executable code when executed by the processor causes the processor to:
determine a first bounding box based on a previous frame;
determine a second bounding box based on a current frame, wherein the current frame is temporally subsequent to the previous frame;
estimate the camera pose by minimizing a weighted sum of a camera pose function and a constraint function, wherein the camera pose function tracks a position and an orientation of the camera in time, and wherein the constraint function is based on coordinates of the first bounding box and coordinates of the second bounding box; and
use at least the camera pose for navigating the vehicle.
10. The apparatus of claim 9 , wherein the constraint function is further based on bounding boxes corresponding to each of a plurality of cars being tracked with a high level of confidence.
11. The apparatus of claim 9 , wherein determining the first and second bounding boxes is based on an initial estimate of the camera pose.
12. The apparatus of claim 11 , further comprising:
a Global Positioning System (GPS) sensor; and
an Inertial Measurement Unit (IMU),
wherein the initial estimate of the camera pose is based on the GPS sensor or the IMU.
13. The apparatus of claim 9 , wherein a weight of the constraint function in the weighted sum is based on previous estimates of the camera pose.
14. The apparatus of claim 9 , wherein the first and second bounding boxes correspond to a position of a rear of a car that is in front of the vehicle.
15. The apparatus of claim 14 , wherein the first bounding box is based on predicting the location of the rear of the car using the previous frame, and wherein the second bounding box is based on detecting the location of the rear of the car using the current frame.
16. A non-transitory computer-readable medium storing a program causing a computer to execute a process, the process comprising:
determining a first bounding box based on a previous frame;
determining a second bounding box based on a current frame, wherein the current frame is temporally subsequent to the previous frame;
estimating the camera pose by minimizing a weighted sum of a camera pose function and a constraint function, wherein the camera pose function tracks a position and an orientation of the camera in time, and wherein the constraint function is based on coordinates of the first bounding box and coordinates of the second bounding box; and
using at least the camera pose for navigating the vehicle.
17. The non-transitory computer-readable medium of claim 16 , wherein the constraint function is further based on bounding boxes corresponding to each of a plurality of cars being tracked with a high level of confidence.
18. The non-transitory computer-readable medium of claim 16 , wherein the first and second bounding boxes correspond to a position of a rear of a car that is in front of the vehicle.
19. The non-transitory computer-readable medium of claim 18 , wherein the first bounding box is based on predicting the location of the rear of the car using the previous frame, and wherein the second bounding box is based on detecting the location of the rear of the car using the current frame.
20. The non-transitory computer-readable medium of claim 16 , wherein a weight of the constraint function in the weighted sum is based on previous estimates of the camera pose.