Lidar Point Cloud Registration for Low-Drift Ego Motion Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ego vehicle motion estimation methods using GPS, IMU, and lidar sensors are prone to errors, drift, and are costly, with lidar-based approaches having low accuracy and being susceptible to drift.
Innovation Solution
A motion monitor system utilizing a lidar device and controller for noncausal and causal registration of lidar point clouds to generate accurate motion sets, including noncausal registration using future scans and causal registration using past scans, with techniques like singular value decomposition and least mean squares to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS/IMU systems are used for ego vehicle motion estimation, then measurement capability is provided, but cost increases and reliability decreases due to errors and drift
Solution Approach 1:
The patent combines multiple lidar scans with GPS/IMU data into a unified motion estimation framework. By aggregating lidar point clouds from multiple scans and registering them together, the system creates a more reliable motion estimate that compensates for individual sensor drift and errors, thereby improving overall reliability while maintaining measurement capability.
Solution Approach 2:
The system implements feedback by using registered lidar point clouds to continuously correct and update motion estimates. The registration process compares current scans with previously registered scans, detecting deviations and correcting transform matrices accordingly. This closed-loop feedback mechanism reduces accumulated drift and improves long-term reliability of motion estimation.
2Ease of manufacture
If lidar-based scan-to-scan registration is used for motion estimation, then cost is reduced, but measurement precision and reliability decrease due to low accuracy and drift
Solution Approach 1:
The patent performs preliminary aggregation of multiple lidar scans before final motion estimation. By pre-registering and aggregating point clouds from multiple scans into a reference frame, the system creates a more robust baseline for comparison. This preliminary action accumulates sufficient geometric features that improve subsequent registration accuracy, thereby enhancing measurement precision while maintaining cost-effectiveness.
Solution Approach 2:
The system uses excessive action by performing multiple registration iterations and aggregating more scans than the minimum required. The patent repeatedly refines transform matrices through multiple registration passes and accumulates a larger number of point clouds than strictly necessary, which improves measurement precision at the cost of increased computational effort, thereby resolving the accuracy limitation of basic lidar registration.
3Measurement precision
If noncausal registration using future scans is implemented, then measurement precision improves, but complexity of operation increases
Solution Approach 1:
The patent performs preliminary aggregation of all available scans into a reference frame before final motion estimation. By pre-processing and organizing point clouds in advance, the system simplifies the actual registration operation. The complex task of handling future scans is broken down into manageable preprocessing steps, making the overall process easier to operate while maintaining high precision through the use of noncausal registration.
4Measurement precision
If multiple scans are aggregated to improve accuracy, then measurement precision increases, but loss of time increases due to processing duration
Solution Approach 1:
The patent segments the large aggregation task into manageable parts by processing scans in batches and using hierarchical registration strategies. Instead of aggregating all scans simultaneously, the system divides them into groups, registers subsets, and progressively builds up the final motion estimate. This segmentation reduces memory requirements and processing time while maintaining the accuracy benefits of multiple-scan aggregation.
Data Source
AI summary
A motion monitor includes a lidar device, a controller, and a circuit. The lidar device is configured to perform scans of an environment proximate the ego vehicle to generate lidar point clouds related to one or more objects in the environment. The controller is configured to transform the lidar point clouds with location transform matrices to generate location point sets, aggregate the location point sets in a duration of interest to generate aggregated point sets, register the location point sets to the aggregated point sets to generate correction transform matrices, update the location transform matrices with the correction transform matrices to generate updated location transform matrices, and generate motion sets based on the updated location transform matrices. The circuit is configured to receive the motion sets.