Map generation using two sources of sensor data
Examples disclosed herein may involve a computing system that is operable to (i) receive first data of one or more geographical environments from a first type of localization sensor, (ii) receive second data of the one or more geographical environments from a second type of localization sensor, (iii) determine constraints from the first data and the second data, (iv) determine shared pose data associated with both of the first data and the second data using the constraints determined from both the first data and the second data by determining one or more sequences of common poses between respective poses generated from each of the first and second data, wherein the shared pose data provides a common coordinate frame for the first data and the second data, and (v) generate a map of the one or more geographical environments using the determined shared pose data.
1. A computer-implemented method comprising:
receiving first data of one or more geographical environments from comprising first sensor data captured by a first type of localization sensor associated with a vehicle;
receiving second data of the one or more geographical environments from comprising second sensor data captured by a second type of localization sensor associated with the vehicle;
generating (i) a first set of submaps from the first data and (ii) a second set of submaps from the second data;
determining (i) a first set of constraints from the first data, wherein the first set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the first set of submaps, and (ii) a second set of constraints from the second data, wherein the second set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the second set of submaps;
merging the first set of constraints determined from the first data with the second set of constraints determined from the second data to form a combined set of constraints; and
applying a combined optimization process to the first set of submaps, the second set of submaps, and the combined set of constraints, wherein the combined optimization process functions to:
determine a single combined trajectory of the vehicle that minimizes an overall error relative to the combined set of constraints;
generate a first map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a first set of vehicles that are installed with the first type of localization sensor but not the second type of localization sensor; and
generate a second map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a second set of vehicles that are installed with the second type of localization sensor but not the first type of localization sensor.
2. The computer-implemented method of claim 1 , wherein the first map comprises a first three-dimensional representation of geographical environments that is used to localize the first set of vehicles and the second map comprises a second three-dimensional representation of geographical environments that is used to localize the second set of vehicles.
3. The computer-implemented method of claim 1 , wherein the first set of submaps is generated using a simultaneous localization and mapping method.
4. The computer-implemented method of claim 3 , wherein the second set of submaps is generated using feature detection.
5. The computer-implemented method of claim 4 , wherein the second set of submaps is refined using a structure from motion method.
6. The computer-implemented method of claim 3 , wherein the first set of submaps is refined using a loop closure technique.
7. The computer-implemented method of claim 1 , wherein the first data and second data are correlated using temporal data to determine one or more relationships between the first and second types of localization sensors.
8. The method of claim 7 , wherein output from the first type of localization sensor and second type of localization sensor are synchronized.
9. The computer-implemented method of claim 1 , wherein each of the first and second types of localization sensors comprises one or more of: a Light Detection and Ranging (LiDAR) sensor, a Radio Detection and Ranging (Radar) sensor, a Sound Navigation and Ranging (Sonar) sensor, an Inertial Navigation System, a Global Positioning System, an Inertial Measurement Unit, or an image sensor.
10. The computer-implemented method of claim 1 , wherein each of the first and second data comprises any one or more of: depth information; point cloud data; or image data.
11. The computer-implemented method of claim 1 , wherein the first data further comprises sensor data captured by one or more additional sensors corresponding to the first type of localization sensor.
12. The computer-implemented method of claim 1 , wherein the second data further comprises sensor data captured by one or more additional sensors corresponding to the second type of localization sensor.
13. The computer-implemented method of claim 1 , wherein the combined set of constraints includes constraints generated from one or more of inertial measurement unit (IMU) data or inertial navigation system (INS) data.
14. The computer-implemented method of claim 1 , wherein the combined set of constraints includes both (i) constraints generated using a loop closure technique and (ii) constraints generated using a structure from motion method.
15. The computer-implemented method of claim 1 , wherein the combined set of constraints includes constraints that relate relative poses of the vehicle between different positions of the vehicle at different times over the vehicle's trajectory.
16. The computer-implemented method of claim 1 , wherein the first map and the second map are correlated to a same coordinate frame.
17. A non-transitory computer-readable medium comprising program instructions stored thereon that, when executed by at least one processor of a computing system, cause the computing system to perform functions comprising:
receiving first data of one or more geographical environments comprising first sensor data captured by a first type of localization sensor associated with a vehicle;
receiving second data of the one or more geographical environments comprising second sensor data captured by a second type of localization sensor associated with the vehicle;
generating (i) a first set of submaps from the first data and (ii) a second set of submaps from the second data;
determining (i) a first set of constraints from the first data, wherein the first set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the first set of submaps, and (ii) a second set of constraints from the second data, wherein the second set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the second set of submaps;
merging the first set of constraints determined from the first data with the second set of constraints determined from the second data to form a combined set of constraints; and
applying a combined optimization process to the first set of submaps, the second set of submaps, and the combined set of constraints, wherein the combined optimization process functions to:
determine a single combined trajectory of the vehicle that minimizes an overall error relative to the combined set of constraints;
generate a first map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a first set of vehicles that are installed with the first type of localization sensor but not the second type of localization sensor; and
generate a second map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a second set of vehicles that are installed with the second type of localization sensor but not the first type of localization sensor.
18. The non-transitory computer-readable medium of claim 17 , wherein:
the first set of submaps is generated using a simultaneous localization and mapping method;
the second set of submaps is generated using feature detection;
the second set of submaps is refined using a structure from motion method; and
the first set of submaps is refined using a loop closure technique.
19. A computing system comprising:
at least one processor, a non-transitory computer-readable medium; and
program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor such that the computing system is configured to perform functions comprising:
receiving first data of one or more geographical environments comprising first sensor data captured by a first type of localization sensor associated with a vehicle;
receiving second data of the one or more geographical environments comprising second sensor data captured by a second type of localization sensor associated with the vehicle;
generating (i) a first set of submaps from the first data and (ii) a second set of submaps from the second data;
determining (i) a first set of constraints from the first data, wherein the first set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the first set of submaps, and (ii) a second set of constraints from the second data, wherein the second set of constraints comprises one or both of intra-submap constraints or inter-submap constraints for the second set of submaps;
merging the first set of constraints determined from the first data with the second set of constraints determined from the second data to form a combined set of constraints; and
applying a combined optimization process to the first set of submaps, the second set of submaps, and the combined set of constraints, wherein the combined optimization process functions to:
determine a single combined trajectory of the vehicle that minimizes an overall error relative to the combined set of constraints;
generate a first map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a first set of vehicles that are installed with the first type of localization sensor but not the second type of localization sensor; and
generate a second map that is (i) built from both the first data and the second data and (ii) thereafter utilized to localize a second set of vehicles that are installed with the second type of localization sensor but not the first type of localization sensor.
20. The computing system of claim 19 , wherein the first map comprises a first three-dimensional representation of geographical environments that is used to localize the first set of vehicles and the second map comprises a second three-dimensional representation of geographical environments that is used to localize the second set of vehicles.