Variable environment high integrity registration transformation system and related method

• US 9,513,130 B1
• Filed: 09/24/2014
• Issued: 12/06/2016
• Est. Priority Date: 09/24/2014
• Status: Active Grant

• Current Assignee: Rockwell Collins, Inc. (Raytheon Technologies Corporation d/b/a RTX)
• Original Assignee: Rockwell Collins, Inc. (Raytheon Technologies Corporation d/b/a RTX)
• Inventors: Kirchner, William T., Xu, Yingyue, Boggs, Christopher M., Allen, Keith M., Kelsey, Gregory J.
• Primary Examiner(s): Olszewski, John R
• Assistant Examiner(s): Foster, Gerrad A
• Application Number:
  US14/495,628
• Time in Patent Office: 804 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G01C 21/1652 with ranging devices, e.g. LIDAR or RADAR
• CPC Class Codes: G05D 1/042 specially adapted for aircraft
• CPC Class Codes: G05D 1/0646 to follow the profile of undulating ground
• CPC Class Codes: G05D 1/0808 specially adapted for aircraft

Abstract

A registration solution between two or more sets of point cloud data is provided to register a first target set of point cloud data with a second source set of point cloud data. Through error propagation, degrees of freedom masking and transformation filtering, a system may provide an intelligent method of enabling and disabling degrees of freedom in a real-time registration application. A system may adapt to varying environments (feature-rich and featureless) ensuring a high-integrity registration estimate through providing estimates of the observable parameters. By optionally and iteratively masking zero or more degrees of freedom from the registration estimate, a system may locate desirable data to use in a registration estimate. Then, the system may blend updates from the registration estimates to achieve the best available registration solution applied to the original second source set as well as additional sets of point cloud data.

Claims

1. A system for iterative closest point registration to accurately determine a position of a vehicle in a substantially featureless environment, comprising:

a memory onboard a vehicle, the memory configured at least for storing a first set of point cloud data, a persistent estimate and computer executable program code;

a sensor onboard the vehicle configured for sensing a second set of point cloud data;

a controller operatively coupled to the memory and the sensor, the controller configured for executing the computer readable program code, the computer readable program code comprising instructions for causing the controller to perform and direct the steps of;

accessing the first set of point cloud data from the memory;

receiving the second set of point cloud data from the sensor;

determining a first transformation estimate via registration of the second set of point cloud data to the first set of point cloud data;

determining a confidence level associated with the first transformation estimate based on an error propagation, the error propagation resulting from measurement error of the sensor;

based on the confidence level associated with the first transformation estimate being below a predetermined value due to a reduced number of features accurately sensed by the sensor, determining an appropriate number of degrees of freedom to use to determine at least one second transformation estimate, the appropriate number being one of;

six degrees of freedom and less than six degrees of freedom;

updating the persistent estimate using the at one second transformation estimate with the appropriate number of degrees of freedom; and

determining the position of the vehicle using the registered set of point cloud data according to the updated persistent estimate and the at least one second transformation estimate, the position for presentation to a display or an additional processing system of the vehicle.

2. The system of claim 1, wherein the position is a second position, and wherein the controller is further configured to perform or direct the steps of:

determining a confidence level associated with the persistent estimate;

determining a confidence level associated with the at least one second transformation estimate;

based on the confidence level associated with the persistent estimate being above the predetermined value, determining a first position of the vehicle using the registered set of point cloud data according to the persistent estimate and the first transformation estimate, the first position for presentation to the display or for the additional processing system of the vehicle;

updating the confidence level associated with the persistent estimate based on the confidence level associated with the at least one second transformation estimate; and

based on the confidence level associated with the updated persistent estimate being above the predetermined value, performing an initial registration of an additional set of point cloud data received from the sensor via registration of the additional set of point cloud data with the first set of point cloud data or with another set of point cloud data from memory.

3. The system of claim 1, wherein the first set of point cloud data represents a trusted measurement of a terrain, and wherein the second set of point cloud data is received from an onboard sensor, the onboard sensor includes a combination of at least:

a light detection and ranging sensor, a radar sensor, a long-wave infrared sensor, an inertial measurement unit, a global navigation satellite system sensor, a visible light sensor, and a radio frequency sensor.

4. The system of claim 1, wherein the second set of point cloud data represents an onboard sensor measurement of a terrain proximal with the vehicle, wherein proximal is based upon a range of the onboard sensor performing the measurement, and wherein the measurement error of the sensor is a result of an inaccurately measured feature or one or more features being eliminated from a field of view of the sensor.

5. The system of claim 1, wherein the first transformation estimate is a six degree of freedom registration estimate and the at least one second transformation estimate is one of a six degree of freedom transformation estimate and a less than six degree of freedom transformation estimate, and wherein the persistent estimate is vector or matrix solution to a linearized representation of a non-linear function, the non-linear function provided as follows:

6. The system of claim 1, wherein determining an appropriate number of degrees of freedom to use to determine at least one second transformation estimate further includes masking at least one degree of freedom based on computed confidence levels.

7. The system of claim 1, wherein updating the persistent estimate further includes updating a blended covariance associated with each of one of a series of registration estimates.

8. The system of claim 1, wherein the confidence level associated with the persistent estimate is filtered for a accuracy, wherein the confidence level associated with one of the first transformation estimate and the at least one second transformation estimate further comprises an accuracy level associated with a second absolute position of the vehicle determined using a first absolute position of the vehicle and a relative position of the vehicle, the relative position being relative to the terrain proximal with the vehicle, and wherein proximal is based upon a range of the onboard sensor utilized in performing the measurement.

9. The system of claim 1, wherein the position is used for localization of the vehicle, updating a 3D environment model related to the vehicle, assessing proximity hazards to a crew of the vehicle, or path planning for the vehicle.

10. A method for iterative closest point (ICP) registration to determine a position of a vehicle in a substantially featureless environment, comprising:

accessing a first set of point cloud data from a memory of a vehicle;

receiving a second set of point cloud data from a sensor of the vehicle;

determining a first transformation estimate via registration of the second set of point cloud data to the first set of point cloud data;

determining a confidence level associated with the first transformation estimate based on an error propagation, the error propagation resulting from measurement error of the sensor;

based on the confidence level associated with the first transformation estimate being below the predetermined value due to a reduced number of features accurately sensed by the sensor, determining an appropriate number of degrees of freedom to use to determine at least one second transformation estimate, the appropriate number being one of;

six degrees of freedom and less than six degrees of freedom;

determining the at least one second transformation estimate and an associated confidence level based on the appropriate number of degrees of freedom;

updating the persistent estimate using the at least one second transformation estimate, the updating based on the confidence level associated with the at least one second transformation estimate;

and

determining the position of the vehicle using the registered set of point cloud data according to the updated persistent estimate and the at least one second transformation estimate and presenting the position to a display of the vehicle.

11. The method of claim 10, wherein the position is a second position, the method further comprising:

determining a confidence level associated with the persistent estimate;

determining a confidence level associated with the at least one second transformation estimate;

based on the confidence level associated with the persistent estimate being above the predetermined value, determining a first position of the vehicle using the registered set of point cloud data according to the persistent estimate and the first transformation estimate, the first position for presentation to the display or for the additional processing system of the vehicle;

updating the confidence level associated with the persistent estimate based on the confidence level associated with the at least one second transformation estimate; and

based on the confidence level associated with the updated persistent estimate being above the predetermined value, performing an initial registration of an additional set of point cloud data received from the sensor via registration of the additional set of point cloud data with the first set of point cloud data or with another set of point cloud data from memory.

12. The method of claim 10, wherein the first set of point cloud data represents a trusted measurement of a terrain, and wherein the second set of point cloud data is received from an onboard sensor, the onboard sensor is a combination of at least:

a light detection and ranging sensor, a radar sensor, a long-wave infrared sensor, an inertial measurement unit, a global navigation satellite system sensor, a visible light sensor, and a radio frequency sensor.

13. The method of claim 10, wherein the second set of point cloud data represents an onboard sensor measurement of a terrain proximal with the vehicle, wherein proximal is based upon a range of the onboard sensor utilized in performing the measurement.

14. The method of claim 10, wherein first transformation estimate is a six degree of freedom registration estimate and the at least one second transformation estimate is a less than six degree of freedom transformation estimate.

15. The method of claim 10, wherein determining an appropriate number of degrees of freedom to use to determine the at least one second transformation estimate further includes masking at least one degree of freedom based on computed confidence levels when the appropriate number is determined to be less than six degrees of freedom, and wherein determining an appropriate number of degrees of freedom to use to determine the at least one second transformation estimate further includes weighting the degrees of freedom and Kalman filtering when the appropriate number is determined to be six degrees of freedom.

16. The method of claim 10, wherein updating the persistent estimate further includes updating a blended covariance matrix associated with each of one of a series of registration estimates.

17. The method of claim 10, wherein the confidence level associated with one of the first transformation estimate and the at least one second transformation estimate further comprises an accuracy level associated with a position of the vehicle relative to the terrain proximal with the vehicle, wherein proximal is based upon a range of the onboard sensor utilized in performing the measurement.

18. The system of claim 11, wherein the confidence level associated with the persistent estimate is Kalman filtered for accuracy.

19. A system for iterative closest point (ICP) registration, comprising:

a memory onboard a mobile platform, the memory configured to store a first set of point cloud data, a persistent estimate and computer readable program code;

a sensor onboard the mobile platform configured for sensing a second set of point cloud data;

a controller operatively coupled to the memory and the sensor, the controller configured for executing the computer readable program code, the computer readable program code comprising instructions for causing the controller to perform and direct the steps of;

performing two or more ICP registration iterations, wherein a first ICP registration iteration comprises;

accessing the first set of point cloud data from the memory;

receiving the second set of point cloud data from the sensor;

determining a first transformation estimate via registration of the second set of point cloud data to the first set of point cloud data;

determining a confidence level associated with the first transformation estimate based on an error propagation, the error propagation resulting from measurement error of the sensor;

based on the confidence level associated with the first transformation estimate being below a predetermined value, determining an appropriate number of degrees of freedom to use to determine at least one second transformation estimate, the appropriate number being one of;

six degrees of freedom and less than six degrees of freedom,

wherein a second ICP registration iteration comprises;

updating the persistent estimate using the at one second transformation estimate with the appropriate number of degrees of freedom, wherein updating the persistent estimate includes masking a degree of freedom or giving less weight to the degree of freedom; and

determining an absolute position of the mobile platform using the registered set of point cloud data according to the updated persistent estimate and the at least one second transformation estimate, the absolute position for presentation to a display or an additional processing system of the mobile platform.

20. The system of claim 19,

wherein the second ICP registration iteration further comprises receiving a third set of point cloud data from the sensor, the at least one second transformation estimate based on a registration of the third set of point cloud data with the first set of point cloud data,

wherein the mobile platform comprises an aircraft,

wherein the persistent estimate is based on a linearized representation of a non-linear function solved using Kalman filtering techniques,

wherein solving using Kalman filtering techniques determines a vector or matrix of values at which a difference function is minimized, the difference function being a difference between the second set of point cloud data and the first set of point cloud data, the second set of point cloud data having received a transformation prior to taking the difference between the sets of point cloud data, and

wherein the absolute position of the aircraft is based on the vector or the matrix of values.