Variable environment high integrity registration transformation system and related method
First Claim
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.
1 Assignment
0 Petitions
Accused Products
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.
6 Citations
20 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. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
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. - View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18)
-
-
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. - View Dependent Claims (20)
-
Specification