LIDAR Segment Merging for Autonomous Driving Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inaccurate object recognition in autonomous driving systems due to over-segmentation of LIDAR data, leading to unreliable neighboring object information and deteriorated route prediction and generation accuracy.
Innovation Solution
A LIDAR data-based object recognition apparatus and segment merging method that generates segments, selects a target segment, determines mergeability based on attribute information, and merges segments to improve accuracy and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If segmentation is performed on LIDAR data to group points from the same object, then object information can be extracted, but over-segmentation occurs causing a single object to be divided into multiple segments
Solution Approach 1:
The patent applies merging by combining multiple over-segmented segments that belong to the same object into a single unified segment. The segment merging unit identifies segments corresponding to the same object and merges them based on spatial proximity and attribute similarity, reducing the number of segments from multiple to one while preserving accurate object information.
2Reliability
If segments are merged to reduce over-segmentation, then object information accuracy improves, but system complexity increases due to additional processing requirements
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing attribute information (such as average reflectivity, point density, and spatial characteristics) for each segment before merging. This preliminary preparation allows the merging process to efficiently compare segments using pre-computed attributes rather than analyzing raw point cloud data, thereby reducing the computational complexity of the merging operation while maintaining reliable object recognition.
3Manufacturing precision
If attribute information is used to determine mergeability, then merging accuracy improves, but processing time increases due to additional attribute calculations
Solution Approach 1:
The patent applies local quality by calculating attribute information selectively for specific regions or segments that are candidates for merging, rather than uniformly processing all segments. The system identifies segments with similar spatial locations and computes their attributes only when mergeability is suspected, thereby reducing overall processing time while maintaining high merging precision for relevant segments.
Data Source
AI summary
A LIDAR data based object recognition apparatus merges segments over-segmented in a LIDAR data based segmentation process. The apparatus includes a segment generator generating a plurality of segments by grouping points acquired from a LIDAR sensor. A target segment selector selects a target segment that is a base for merging from the plurality of segments and a segment merging determination unit checks whether segments other than the target segment are mergeable segments and determines whether to merge the target segment and the mergeable segments based on attribute information of the target segment and the mergeable segments. A segment merger merges the target segment and the mergeable segments and outputs a merged segment.