Monocular visual-inertial alignment for scaled distance estimation on mobile devices
Methods, techniques, apparatus, and algorithms are described for robustly measuring real-world distances using any mobile device equipped with an accelerometer and monocular camera. A general software implementation processes 2D video, precisely tracking points of interest across frames to estimate the unsealed trajectory of the device, which is used to correct the device's inertially derived trajectory. The visual and inertial trajectories are then aligned in scale space to estimate the physical distance travelled by the device and the true distance between the visually tracked points.
1. A method, using a mobile device equipped with a camera and an inertial measurement unit (IMU), for measuring real-world distance between two or more points in an image, the method comprising:
aligning estimated visual and inertial trajectories;
converting the visual and inertial trajectory signals into velocity signals; and
scale aligning the velocity signals by minimizing their Euclidean distance.
2. The method of claim 1 , wherein a plug-in image-based feature tracker is used to estimate visual trajectories.
3. The method of claim 2 , wherein the feature tracker is specific to known object landmarks or a generic feature tracker.
4. The method of claim 1 , wherein the mobile device follows a cyclical motion pattern during data capture.
5. The method of claim 4 , wherein the motion has additional constraints on device rotation and velocity that are actively monitored and corrected by software.
6. The method of claim 1 , wherein image distortion is corrected prior to feature tracking.
7. The method of claim 1 , wherein the visual and inertial trajectory signals are preprocessed through standard filtering and time alignment.
8. The method of claim 1 , wherein the inertial trajectory signal drift is corrected via zero velocity detection in the visual trajectory signal.
9. A processor system for measuring real-world distance using a mobile device equipped with a camera and an inertial measurement unit (IMU), the system comprising:
a processor; and
a memory unit in communication with the processor via a communication infrastructure and configured to store processor-readable instructions;
wherein, when executed by the processor, the processor-readable instructions cause the processor to:
align estimated visual and inertial trajectories;
convert the visual and inertial trajectory signals into velocity signals; and
scale align the velocity signals by minimizing their Euclidean distance.
10. The system of claim 9 , wherein a plug-in image-based feature tracker is used to estimate visual trajectories.
11. The system of claim 10 , wherein the feature tracker is specific to known object landmarks or a generic feature tracker.
12. The system of claim 9 , wherein the mobile device follows a cyclical motion pattern during data capture.
13. The system of claim 12 , wherein the motion has additional constraints on device rotation and velocity that are actively monitored and corrected by software.
14. The system of claim 9 , wherein image distortion is corrected prior to feature tracking.
15. The system of claim 9 , wherein the visual and inertial trajectory signals are preprocessed through standard filtering and time alignment.
16. The system of claim 9 , wherein the inertial trajectory signal drift is corrected via zero velocity detection in the visual trajectory signal.