LiDAR Data Processing via Subset Matching and Offset Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LiDAR data processing techniques are inefficient and inaccurate in identifying the same objects across different LiDAR datasets captured at different times, leading to computationally intensive and time-consuming processes, and are not well-suited for object detection due to data sparsity and offset issues, resulting in unreliable analysis and frequent costly actions in real-world applications.
Innovation Solution
A method and system for processing LiDAR data by obtaining and dividing datasets into subsets, matching them to minimize offsets, detecting objects, determining average offsets, creating links between objects within a threshold distance, and evaluating link validity based on a growth criterion, enabling efficient and accurate object identification and analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional point cloud classification techniques are used to identify objects in LiDAR data, then object detection can be performed, but the process becomes computationally intensive and time-consuming due to the huge amount of data generated
Solution Approach 1:
The patent divides the LiDAR dataset into multiple subsets based on spatial regions or object types. Each subset is processed independently through classification, and results are integrated to form the complete object identification. This segmentation reduces the computational burden on any single processing unit while maintaining comprehensive coverage of the entire dataset.
2Reliability
If LiDAR scans are performed frequently to ensure reliable detection, then object detection reliability improves, but business costs and operational inconvenience increase
Solution Approach 1:
The patent performs preliminary processing of LiDAR data including noise filtering, point cloud normalization, and feature extraction before main analysis. By preparing the data in advance with these preliminary steps, the system achieves more reliable object detection from fewer scans, reducing the frequency required and thereby lowering operational costs and time loss.
3Area of stationary object
If LiDAR data is collected from different routes and heights, then comprehensive environmental coverage is achieved, but offset issues and perspective variations reduce object detection accuracy
Solution Approach 1:
The patent applies coordinate transformation and registration techniques to normalize LiDAR data from different routes and heights into a common reference frame. By transforming all data points to equivalent positional relationships relative to the environment being mapped, the system eliminates offset issues and perspective variations, enabling accurate object detection across comprehensively collected data.
4Productivity
If point cloud data is sparsely sampled, then data collection speed increases, but available data per object becomes limited reducing detection accuracy
Solution Approach 1:
The patent combines multiple LiDAR scans and supplementary data sources (such as imagery or sensor data) to create denser, more complete point clouds for each object. By merging data from multiple acquisitions and sources, the system compensates for sparsity in individual scans, providing sufficient data volume per object while maintaining efficient collection speeds through parallel processing.
Data Source
AI summary
A method including: obtaining first LiDAR dataset and second LiDAR dataset of environment from LiDAR database, first LiDAR dataset and second LiDAR dataset being captured at first time period and second time period, respectively, second time period being later than first time period; dividing first and second LiDAR datasets into first and second LiDAR subsets; matching given first LiDAR subset with given second LiDAR subset; detecting first objects and second objects in given first and second LiDAR subsets; determining average offset between locations of first objects and locations of second objects; creating given link between given first object and at least one second object, said second object lying within predefined threshold distance from given first object in direction of average offset; and evaluating validity of given link, based upon whether or not given link satisfies growth criterion, given first object is associated with at most one valid link.