Static Obstacle Identification Using Laser Point Cloud Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying static obstacles in autonomous driving environments suffer from low accuracy due to noise in speed data, leading to incorrect classification of obstacles as static or moving.
Innovation Solution
A method and apparatus that generate determining information by comparing historical and current laser point cloud data of obstacles, including similarity, matching degree, and overlap ratio, to determine if an obstacle is static, considering noise type motion characteristics and historical motion patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion estimation filtering algorithm is used to calculate speed and identify obstacles below threshold as static obstacles, then the identification process is simple and fast, but the accuracy is low due to noise data in speed measurements
Solution Approach 1:
The patent segments the identification process into multiple independent evaluation dimensions: motion characteristic similarity, appearance characteristic matching degree, and overlap ratio. Each dimension is calculated and evaluated separately, then integrated to form a comprehensive static obstacle identification decision, thereby improving accuracy without excessive complexity
Solution Approach 2:
The patent transitions from single-dimensional speed threshold judgment to multi-dimensional evaluation by introducing motion characteristic similarity, appearance characteristic matching degree, and overlap ratio as additional evaluation dimensions. This dimensional expansion enables more accurate static obstacle identification by considering multiple aspects simultaneously
2Reliability
If multiple determination information including similarity, matching degree, and overlap ratio are calculated, then the identification accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent implements partial action by selectively applying multiple determination information calculations only when historical laser point cloud sequences are available for the obstacle. This approach improves reliability through multi-factor evaluation while avoiding unnecessary computational complexity when historical data is insufficient
Solution Approach 2:
The patent performs preliminary action by pre-calculating and storing motion characteristics and appearance characteristics during the obstacle tracking process. These pre-computed features are then reused in the static obstacle identification process, improving reliability without repeating redundant calculations
Data Source
AI summary
The disclosure discloses a method and apparatus for identifying a static obstacle. An embodiment of the method includes: determining, based on determining information corresponding to a historical laser point cloud sequence of an obstacle, a detected laser point cloud of the obstacle in a current laser point cloud frame belonging to the historical laser point cloud sequence of the obstacle, whether a given obstacle is a static obstacle. The determining information includes: a similarity between a historical motion characteristic of the given obstacle and a noise type motion characteristic, a matching degree between an appearance characteristic of the detected laser point cloud of the obstacle and an appearance characteristic of the historical laser point cloud of the obstacle, and an overlap ratio between the detected laser point cloud of the obstacle and the historical laser point cloud of the obstacle.