LiDAR Curb Detection via Point Cloud Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional LiDAR-based curb detection methods face challenges such as noise interference from objects like vehicles and bushes, occlusions, and the diversity of curb types, making accurate and robust detection difficult, especially when curbs are far away or obscured.
Innovation Solution
The method employs a LiDAR sensor to generate point-cloud data, which is pre-processed and used for robust point clustering by selecting seeds, growing clusters based on criteria like height and curvature, and computing the concave hull to identify the road surface boundary, thereby detecting the curb more accurately and robustly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional LiDAR-based curb detection methods are used, then the detection process is simple, but the detection accuracy deteriorates due to noise interference from objects like vehicles and bushes
Solution Approach 1:
The patent segments the point cloud data into multiple clusters based on spatial proximity and geometric characteristics. Each cluster is processed independently to identify curb segments, which helps filter out noise from objects like vehicles and bushes by isolating them from the road surface points. This segmentation approach enables accurate curb detection even in the presence of interfering objects.
Solution Approach 2:
The patent extracts and removes outlier points that do not conform to the expected road surface geometry. By identifying and excluding these anomalous points (which correspond to noise from vehicles, bushes, and other objects), the method purifies the point cloud data and enhances curb detection accuracy without being misled by harmful factors.
2Reliability
If conventional LiDAR-based curb detection methods are used, then the system is simple, but the detection reliability deteriorates due to occlusions
Solution Approach 1:
The patent divides the point cloud into multiple spatial clusters and processes each cluster independently. This segmentation allows the system to detect curbs in different segments even when some segments are occluded by objects like vehicles or bushes, maintaining overall detection reliability through parallel processing of multiple regions.
Solution Approach 2:
The patent performs preliminary clustering and geometric analysis on the point cloud data before final curb detection. By pre-organizing the data into clusters and identifying geometric patterns in advance, the system can better handle occlusions and maintain reliable detection even when parts of the curb are obscured by interfering objects.
3Measurement precision
If conventional LiDAR-based curb detection methods are used, then the processing is simple, but the detection accuracy deteriorates when curbs are far away or obscured
Solution Approach 1:
The patent segments the distant point cloud data into multiple spatial clusters, allowing the system to process and analyze curb characteristics in smaller, manageable segments. This segmentation enables accurate detection of distant curbs by maintaining geometric relationships and patterns even when points are far from the sensor, overcoming the accuracy degradation that occurs with conventional methods at long distances.
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 allows for accurate and robust curb detection in 3D space, even in the presence of noise and occlusions, by focusing on road surface expansion and combining multiple features for point clustering, enhancing the reliability of autonomous vehicle navigation.
Implementation Method 1
light detection and ranging (LiDAR) sensors may be used to acquire the images based on reflections captured from the surrounding area
Implementation Method 2
acquire the images based on reflections captured from the surrounding area
Data Source
AI summary
Disclosed are devices, systems and methods for road feature detection using light detection and ranging (LiDAR) sensors. One example of a method for road feature detection includes obtaining a point-cloud frame that comprises a description of an intensity of a reflection of beams from an area around a vehicle. The example method further includes creating a plurality of clusters that each include (i) one or more seed points of the point-cloud frame, and (ii) additional points of the point-cloud frame based on a relationship between the additional points and the one or more seed points. The example method further includes identifying a cluster from the plurality of clusters based on a total number of points included in each of the plurality of clusters. The example method further includes detecting a road feature from the cluster.