Radar Track Center Pointing for Ghost Object Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional radar-based object tracking systems for advanced driver assistance systems (ADAS) and autonomous driving face challenges such as ghost detection and inaccurate object estimation due to the use of single-point detection, especially in complex scenarios with multiple objects, leading to difficulties in distinguishing actual targets from ghost detections.
Innovation Solution
An object tracking device that utilizes a center point of a track, comprising a track setter, a crossing point counter, and an object tracker to set and reset tracks based on location and movement information, determining and removing ghost detection points by analyzing crossing points between detection points and tracks, thereby enhancing accuracy in object tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional radar-based object tracking using single point detection is used, then the system is simple to operate, but ghost detection occurs and object tracking accuracy deteriorates
Solution Approach 1:
The patent segments the tracking process into multiple components: setting initial tracks, extending tracks, determining crossing points, and resetting tracks based on crossing point counts. This segmentation allows the system to handle complex tracking scenarios through modular operations, improving accuracy without overwhelming complexity
Solution Approach 2:
The patent introduces a new dimension of analysis by counting crossing points between detection points and tracks. Instead of relying solely on single-point detection data, the system analyzes the geometric relationship (crossing points) between detection points and track trajectories, adding a spatial dimension to ghost detection elimination
2Measurement precision
If artificial intelligence model is used to remove ghost detection points, then object tracking accuracy improves, but training difficulty and time consumption increase
Solution Approach 1:
The patent replaces expensive and time-consuming AI model training with a simpler, computationally lightweight crossing point counting method. This approach uses basic geometric calculations that can be performed in real-time without extensive training data or model optimization, effectively using simple disposable calculations instead of heavy AI processing
Solution Approach 2:
The patent substitutes the complex AI/machine learning system with a deterministic geometric algorithm. Instead of using neural networks and training models, the system employs mathematical calculations to count crossing points and identify ghost detections, replacing software-intensive AI with efficient mathematical operations
Data Source
AI summary
The disclosure relates to a technology for tracking an object using a center point of a track and provides an object tracking device and method, comprising receiving location information about detection points for an object and setting a first track of the object and a second track extended from the first track based on the location information, determining the number of crossing points between a line connecting the detection point and the radar and any one of the first track or the second track based on movement information about a first track center point, and resetting the first track and the second track based on location information about a detection point in which the number of crossing points is less than N as preset (where N is an integer of two or more) and tracking the object based on any one of the reset first track or second track.