Automobile Radar Target Detection Using Kalman Filter Stabilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radar systems for motor vehicles face challenges in differentiating target objects from noise and distinguishing between moving and stationary objects due to noisy and unstable information, making it difficult to accurately detect and track objects in the vehicle's environment.
Innovation Solution
A method that corrects target point positions based on vehicle movement, groups neighboring points using spatial and power/speed characteristics, and applies a Kalman filter to stabilize position calculations, allowing for the differentiation of target objects and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If radar scanning is used to detect target objects, then detection capability is provided, but the returned information becomes noisy and unstable making it difficult to differentiate target objects from noise
Solution Approach 1:
The patent applies preliminary action by correcting target point positions using previously stored vehicle position information before grouping and analysis. The system stores vehicle positions at multiple time points and uses this historical data to compensate for vehicle movement effects on target point coordinates, thereby stabilizing the target point information before further processing
Solution Approach 2:
The patent implements feedback by continuously comparing target point positions across multiple time points and using the observed patterns to refine detection. The system analyzes target point stability over time and uses this feedback to differentiate between stationary objects (stable positions) and moving objects (changing positions), as well as to filter noise
2Difficulty of detecting and measuring
If radar scanning is used to detect target objects, then detection capability is provided, but it becomes difficult to differentiate between moving and stationary objects
Solution Approach 1:
The patent performs preliminary classification by analyzing target point position stability over time before final object identification. By comparing target point coordinates across multiple time points and determining whether they remain within a threshold distance, the system pre-categorizes targets as stationary or moving, which simplifies subsequent processing
Solution Approach 2:
The patent applies segmentation by dividing the detection process into distinct stages: position correction, grouping analysis, stability verification, and classification. This segmented approach allows the system to handle different aspects of target differentiation separately, improving overall classification accuracy
3Difficulty of detecting and measuring
If radar scanning is used to detect target objects, then detection capability is provided, but it becomes difficult to differentiate between close objects in the environment
Solution Approach 1:
The patent uses segmentation by implementing a grouping step that clusters nearby target points based on spatial proximity. By analyzing the spatial distribution and density of target points, the system can separate closely spaced objects into distinct groups, enabling accurate differentiation even when objects are near each other
Solution Approach 2:
The patent applies dimensionality change by analyzing target points not only in spatial coordinates but also in the time dimension. By examining target point positions across multiple time points and verifying stability patterns, the system adds temporal dimensionality to the analysis, which helps differentiate closely spaced objects that may have different movement characteristics
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for detecting a target object for a motor vehicle. It is characterized in that the method comprises the steps of: - correcting the position of at least one target point relative to a motor vehicle, as a function of the motor vehicle's movement over a determined number of cycles; - starting from at least one target point, forming a first group with neighboring target points according to a first determined characteristic; - verifying whether the first group is homogeneous according to a second determined characteristic; and - calculating the position of a group formed relative to the motor vehicle over the determined number of cycles, each group formed corresponding to a target object.