Autonomous Vehicle Lidar ROI Point Cloud Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In autonomous driving systems, existing technologies face challenges in efficiently detecting and classifying obstacles around the vehicle using sensor data, particularly in determining relevant point cloud data for safe driving, which affects the decision-making and control processes.
Innovation Solution
A method and apparatus that determine a target area around the autonomous vehicle using positioning information, acquire point cloud data from a lidar, and convert it into a world coordinate system, utilizing polygon scan line filling algorithms or grid graph techniques to generate and output relevant point cloud data sets for obstacle detection and analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the lidar detects obstacles in the entire preset area around the autonomous vehicle, then the coverage area is comprehensive, but the processing complexity and time consumption increase significantly
Solution Approach 1:
The patent divides the preset detection area into multiple sub-areas based on the target area (region of interest). The lidar selectively detects obstacles only in these divided sub-areas rather than the entire preset area, reducing the amount of data to be processed while maintaining detection effectiveness in critical regions.
Solution Approach 2:
The patent applies different detection strategies to different spatial regions. High-density detection is applied to the target area (region of interest) where obstacles have significant impact, while lower-density detection is applied to other areas, optimizing the balance between detection quality and processing load.
2Reliability
If the lidar detects and processes all point cloud data in the preset area, then the obstacle detection completeness is high, but the processing time and computational resources increase
Solution Approach 1:
The patent extracts and processes only the point cloud data corresponding to the target area (region of interest) from the complete preset area data. By separating the critical region data from the rest, the system achieves high detection completeness for relevant obstacles while significantly reducing processing time and computational resource consumption.
3Measurement precision
If the system processes point cloud data from the entire preset area, then the obstacle detection accuracy is maintained, but the processor usage and energy consumption increase
Solution Approach 1:
The patent segments the point cloud data processing task by identifying and processing only the subset of data corresponding to the target area. This segmentation maintains detection accuracy for obstacles in the region of interest while reducing overall processor energy consumption by eliminating unnecessary processing of data from non-critical areas.
4Area of stationary object
If the lidar scans the entire surrounding area with radius 60m, then the detection coverage is comprehensive, but the data volume and processing load increase
Solution Approach 1:
The patent extracts only the point cloud data corresponding to the target area (region of interest) from the comprehensive preset area data collected by the lidar. This extraction reduces the data volume significantly while maintaining detection coverage for the most critical areas where obstacles would have the greatest impact on vehicle safety.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy and efficiency of obstacle detection, reducing processor usage and improving driving safety by focusing on the perceived region of interest, thereby facilitating safer autonomous vehicle operations.
Implementation Method 1
The autonomous vehicle includes a lidar. The method includes: acquiring point cloud data of an obstacle in a preset area around the autonomous vehicle by the lidar
Data Source
AI summary
A method and apparatus for outputting information of an autonomous vehicle are provided. The autonomous vehicle comprises a lidar. A specific embodiment of the method comprises: determining a target area around the autonomous vehicle from a target map based on positioning information of the autonomous vehicle; acquiring obstacle point cloud data in a preset area around the autonomous vehicle by the lidar, the preset area including the target area; determining a point cloud data set corresponding to the target area from the point cloud data of the obstacle based on the target area and the point cloud data, and determining the point cloud data set as a target point cloud data set; and outputting the target point cloud data set. The embodiment has reduced the usage rate of vehicle terminal processors, and improved the driving safety of the vehicle.