LiDAR Resolution Control Using Infrared and Event Sensor Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current LiDAR systems in vehicles require slower response times for adjusting resolution configurations, which can delay remedial actions in dynamic driving scenarios, as they rely on perception components for object detection and tracking.
Innovation Solution
Implementing a system where an infrared camera or event-based sensor generates outputs that bypass the perception component, allowing the LiDAR to configure its resolution directly by positioning a higher-resolution region within its field of view based on detected temperature or motion differences, thereby enabling faster adjustment of scanning rates and laser pulse frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the LiDAR relies on the perception component for object detection and tracking to adjust resolution configuration, then the system can perform comprehensive object analysis, but the response time for resolution adjustments increases
Solution Approach 1:
The system segments the object detection pipeline by creating two parallel paths: a fast path using infrared cameras and event-based sensors that bypasses the full perception component to quickly identify objects of interest, and a comprehensive path that uses the full perception stack for detailed analysis. This segmentation allows the LiDAR to receive quick updates about objects of interest without waiting for the complete perception processing pipeline, thereby reducing response time while maintaining detection accuracy.
Solution Approach 2:
The system performs preliminary object detection and tracking using infrared cameras and event-based sensors before the LiDAR needs to adjust its resolution. These sensors continuously monitor the environment and pre-identify objects of interest, so when the LiDAR needs to adjust its resolution configuration, the target objects are already identified and ready for high-resolution scanning. This preliminary action eliminates the waiting time that would otherwise be required for object detection.
2Measurement precision
If the LiDAR uses a higher-resolution region within the field of view for detailed scanning, then the detection precision improves, but the scanning rate decreases
Solution Approach 1:
The LiDAR implements local quality by dividing its field of view into different resolution regions: a higher-resolution region (HRR) for detailed scanning of identified objects of interest, and lower-resolution regions for the surrounding environment. The system dynamically adjusts which areas receive high-resolution scanning based on object importance, allowing concentrated computational resources to be applied only where needed. This enables the system to maintain high detection precision for critical objects while preserving an acceptable overall scanning rate across the entire field of view.
Solution Approach 2:
The system applies partial action by selectively scanning only the higher-resolution region containing objects of interest at high detail levels, while using lower resolution for the remainder of the field of view. This partial application of high-resolution scanning to only the necessary portion of the scene allows the system to achieve sufficient detection precision for safety-critical objects without the excessive computational burden of scanning the entire field at maximum resolution, thereby maintaining an acceptable scanning rate.
3Speed
If the LiDAR adjusts laser pulse frequency and scanning rate dynamically, then the response speed improves, but the system complexity increases
Solution Approach 1:
The system implements feedback by continuously monitoring object distance, velocity, and classification from infrared cameras and event-based sensors, then using this information to dynamically adjust the LiDAR's scanning rate and pulse frequency. When objects are detected at close distances or moving rapidly, the system automatically increases scanning rate and pulse frequency to maintain tracking accuracy. This closed-loop feedback mechanism enables automatic adaptive control without requiring complex manual configuration or intervention.
Solution Approach 2:
The LiDAR system transitions from static, fixed-parameter operation to dynamic, adaptive operation where scanning rate and pulse frequency automatically adjust based on real-time environmental conditions and object characteristics. The system dynamically modifies its operational parameters in response to changing scene complexity, object distance, and motion rates, allowing optimal performance across varying driving conditions without requiring complex manual reconfiguration.
Data Source
AI summary
A computer-implemented method comprises: generating first output using a first sensor of a vehicle comprising an infrared camera or an event-based sensor, the first output indicating a portion of surroundings of the vehicle; providing the first output to a LiDAR of the vehicle having a field of view (FOV); configuring a resolution of the LiDAR based at least in part on the first output; generating a representation of at least part of the surroundings of the vehicle using the LiDAR; providing, to a perception component of the vehicle, second output of a second sensor of the vehicle and third output of the LiDAR, the perception component configured to perform object detection, sensor fusion, and object tracking regarding the second and third outputs, wherein the first output bypasses at least part of the perception component; and performing motion control of the vehicle using a fourth output of the perception component.