Radar Path Estimation Using Stationary Object Curve Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radar-equipped vehicles face challenges in maintaining control targets, especially when a set control target enters a curved road, as image sensors may fail to detect the control target, leading to inaccuracies in estimating the vehicle driving path.
Innovation Solution
A radar control device and method that includes a receiver for detecting objects around the host vehicle, a producer for classifying measurements as moving or stationary objects and performing curve fitting on stationary objects to produce a first path, and an estimator for determining the vehicle driving path based on this path, ensuring accurate control even on curved roads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image sensors are used for smart cruise control, then detection accuracy is improved on straight roads, but detection reliability deteriorates on curved roads
Solution Approach 1:
The patent combines radar sensors and image sensors into a unified detection system. The radar sensor continuously tracks the control target's distance and relative velocity, while the image sensor provides visual confirmation on straight roads. This merging allows the system to maintain detection reliability on curved roads using radar while benefiting from image sensor precision on straight roads.
Solution Approach 2:
The system dynamically changes detection parameters based on road conditions. When curvature is detected, the system switches to radar-based detection with adjusted tracking parameters, while on straight roads it uses image sensor-based detection. This parameter adaptation resolves the contradiction by optimizing detection methods for different operational conditions.
2Measurement precision
If curve fitting is performed on stationary objects, then path estimation accuracy is improved, but system complexity increases
Solution Approach 1:
The patent extracts only the necessary stationary objects (roadside objects like trees, poles, signs) for curve fitting operations, rather than processing all detected objects. By selectively identifying and fitting only relevant stationary objects that define the road boundary, the system achieves accurate path estimation while minimizing computational complexity.
Solution Approach 2:
The detection field is segmented into moving objects (vehicles, pedestrians) and stationary objects (roadside infrastructure). Only the stationary objects are subjected to curve fitting operations for path estimation, while moving objects are handled separately for tracking. This segmentation reduces the complexity of path estimation by limiting curve fitting to a specific subset of objects.
Data Source
AI summary
The disclosure relates to a radar control device and method. Specifically, a radar control device according to the disclosure comprises a receiver receiving reception information obtained by detecting an object around a host vehicle, a producer detecting a measurement based on the reception information, classifying the measurement as a moving object and a stationary object, and performing a curve fitting on the stationary object to produce a first path, and an estimator estimating a vehicle driving path based on the first path.