LiDAR Object Tracking Using Historical Shape Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inaccurate information from LiDAR sensors can lead to erroneous performance of Highway Driving Pilot (HDP) systems, compromising the reliability of object tracking in vehicles.
Innovation Solution
A method and apparatus for tracking objects using LiDAR sensors, which involves determining a reference point for a segment box, checking associations with historical shape information, and generating current tracking box information to accurately update and output the shape of the target object, including clustering point cloud data and selecting the associated segment box.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If LiDAR sensor information is used for HDP system, then object tracking function is enabled, but tracking accuracy deteriorates due to inaccurate sensor data
Solution Approach 1:
The patent implements feedback by accumulating history shape information from previous tracking frames and using it to verify and correct current tracking box information. The system continuously compares current LiDAR data with historical data, allowing the tracking system to self-correct errors and maintain accuracy despite noisy sensor inputs.
Solution Approach 2:
The patent performs preliminary action by pre-accumulating history shape information before it is needed for correction. The system maintains a repository of past object shapes and positions that can be quickly referenced when current sensor data is unreliable, enabling rapid correction without waiting for multiple failed measurements.
2Productivity
If simple LiDAR point cloud data is used, then processing speed is maintained, but tracking accuracy deteriorates
Solution Approach 1:
The patent introduces an intermediary element - the history shape information - that mediates between the simple LiDAR point cloud data and the required tracking accuracy. This historical data acts as a bridge, providing the additional shape context needed for accurate tracking without requiring complex real-time processing of raw point clouds.
Solution Approach 2:
The patent uses copying by creating a simplified representation of the object shape from historical data that can be quickly compared with current sensor input. Instead of processing complex 3D point clouds in real-time, the system copies and compares essential shape features from history, maintaining processing speed while improving accuracy.
3Stability of the object's composition
If tracking box information is updated frequently, then tracking stability is improved, but error propagation increases
Solution Approach 1:
The patent uses feedback to prevent error propagation by continuously verifying tracking box updates against historical shape information. Before accepting new tracking data, the system checks it against accumulated history, allowing it to reject erroneous updates while maintaining stability through validated corrections.
Solution Approach 2:
The patent implements beforehand cushioning by maintaining a buffer of history shape information that cushions against erroneous tracking updates. When current data conflicts with historical patterns, this pre-accumulated information acts as a protective buffer, preventing error propagation while allowing legitimate tracking adjustments.
Data Source
AI summary
An object-tracking method using a LiDAR sensor includes generating current shape information about a current tracking box at a current time from an associated segment box, using history shape information accumulated prior to the current time with respect to a target object that is being tracked, and updating information on a previous tracking box at a time prior to the current time, contained in the history shape information, using the current shape information and the history shape information and determining a previous tracking box having the updated information to be a final output box containing information on the shape of the target object.