LiDAR Road-Curb Detection Using Circular Grid Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting road-curbs, such as those using cameras and LiDAR data, face challenges like illumination dependence and processing speed limitations, especially in complex urban environments where accurate and rapid detection is crucial for autonomous driving.
Innovation Solution
A method utilizing a LiDAR sensor to obtain point data, which is arranged in a circular grid map, allowing for the detection of road-curbs without requiring all point data and enabling rapid and accurate identification regardless of road-curb direction, by determining candidate cells based on height differences and performing line fitting processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera-based road-curb detection is used, then road-curb can be detected during daytime, but detection performance deteriorates under varying illuminance conditions and cannot detect road-curbs in low-light or nighttime conditions
Solution Approach 1:
The patent replaces the camera-based optical detection system with a LiDAR-based laser ranging system. LiDAR uses laser beams to measure distances and detect road-curbs through time-of-flight measurements, completely avoiding dependence on visible light illumination. This substitution enables reliable road-curb detection in all lighting conditions including nighttime and low-light environments, while maintaining detection accuracy independent of illuminance changes.
2Measurement precision
If conventional LiDAR road-curb detection using all point data is used, then detection can be performed without layer information, but processing speed is slow
Solution Approach 1:
The patent divides the point cloud data into multiple horizontal layers based on height differences, creating a layered structure from the LiDAR point data. This segmentation allows the system to process only relevant layers for road-curb detection rather than analyzing all point data, significantly reducing computational complexity and processing time while maintaining accurate road-curb detection capability.
Solution Approach 2:
The patent applies partial action by selectively processing only the necessary horizontal layers that contain road-curb information, rather than processing all LiDAR point data. By identifying and focusing computation on specific height ranges and layers where road-curbs are likely to exist, the system achieves fast processing speeds suitable for real-time autonomous driving applications while maintaining detection accuracy.
3Reliability
If road-curb detection is performed in complex urban environments, then accurate boundary detection is crucial for safety, but detection becomes more difficult due to complex terrain and multiple road-curbs
Solution Approach 1:
The patent segments complex urban environments into multiple horizontal layers based on height information from LiDAR data. This layering approach simplifies the complex terrain by organizing points at different elevations into distinct layers, making it easier to identify road-curb boundaries even in complex urban settings with multiple road-curbs and varying terrain.
Solution Approach 2:
The patent introduces a height dimension to organize and analyze road-curb data, creating horizontal layers at different elevations. This dimensional approach transforms the complex three-dimensional urban environment into a structured layered representation, enabling more reliable road-curb detection by leveraging vertical height differences between road surfaces and curbs.
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 improves the safety and efficiency of autonomous driving by enabling accurate and rapid road-curb detection, suitable for real-time operation, even in environments with unknown layer information.
Implementation Method 1
a light detection and ranging (LiDAR) sensor
Data Source
AI summary
A method of detecting a road-curb that is performed by a road-curb detecting apparatus is provided. The method includes obtaining points around a lidar sensor from the lidar sensor, arranging the points in a plurality of cells into which a circular grid map is divided, and detecting the road-curb based on the points arranged in the plurality of cells.