LiDAR Object Tracking With Variance-Adaptive Kalman Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems using LiDAR are prone to generating imprecise data due to environmental noise, leading to false tracking or loss of tracking targets, which can hinder safe vehicle control.
Innovation Solution
An object detection device that utilizes a Kalman filter to periodically update and adjust the output timing and duration of predicted position and speed values based on the variance of these values, enhancing tracking accuracy by ensuring reliable predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If continuous tracking is performed using Kalman filter with fixed periodic output, then tracking coverage is maintained, but tracking accuracy deteriorates due to noise in noisy environments
Solution Approach 1:
The patent dynamically changes the output period parameter of the Kalman filter based on the variance of predicted values. When variance is high (noisy environment), the output period is extended to filter out noise. When variance is low (clean environment), the output period is shortened to provide timely tracking updates. This adaptive parameter adjustment resolves the contradiction between maintaining tracking coverage and improving tracking accuracy in noisy environments.
Solution Approach 2:
The tracking system transitions from a static fixed-period output mechanism to a dynamic adaptive output mechanism. The output period and tracking duration are dynamically adjusted based on real-time variance assessment of predicted position values. This dynamic adaptation allows the system to optimize tracking accuracy while maintaining continuous tracking coverage under varying environmental conditions.
2Reliability
If tracking duration is extended to maintain continuous tracking, then tracking coverage is improved, but false tracking of noise images increases
Solution Approach 1:
The patent implements a feedback mechanism where the variance of predicted position values is continuously monitored and used to adjust tracking parameters. When variance exceeds a threshold (indicating potential false tracking of noise), the system shortens tracking duration and extends output period to suppress false tracking. When variance is below threshold (indicating reliable tracking), the system extends tracking duration to maintain continuity. This feedback control resolves the contradiction between tracking continuity and false tracking suppression.
Solution Approach 2:
The system takes preliminary anti-action by detecting high variance conditions that precede false tracking of noise images. When variance is high, the system proactively shortens tracking duration and extends output period before false tracking can occur, thereby preventing false tracking of noise images while maintaining overall tracking continuity.
3Speed
If output period is shortened to provide timely tracking updates, then tracking responsiveness is improved, but reliability of predicted values decreases due to noise
Solution Approach 1:
The patent dynamically changes the output period parameter based on variance of predicted values. In noisy environments (high variance), the output period is extended to improve reliability of predicted values. In clean environments (low variance), the output period is shortened to provide timely tracking updates. This adaptive parameter adjustment resolves the contradiction between tracking update speed and predicted value reliability.
Data Source
AI summary
An object detection device according to one aspect includes: a point cloud data acquisition unit configured to acquire point cloud data generated by a laser sensor; an object detection unit configured to periodically acquire position information on a tracking target object, based on the point cloud data; a tracking unit configured to input the position information on the tracking target object to a Kalman filter in a first period and output a predicted value of a position or speed of the tracking target object in a second period shorter than the first period; and a tracking control unit configured to change an output timing of the predicted value from the tracking unit or a tracking duration for the tracking target object, based on a variance of the predicted value.