Fusion Tracking for Accurate Stationary Target Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automotive radars have limited angle recognition performance and accuracy in identifying stationary targets, which hinders effective autonomous driving systems that rely on sensor fusion for real-time environment recognition.
Innovation Solution
A method and device that generate a fusion track based on data from both radar and camera systems to determine the position and contour of stationary targets by collecting and processing radar points, improving accuracy and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar data alone is used for stationary target detection, then the system complexity is low, but the measurement precision and reliability are insufficient due to limited angle recognition performance
Solution Approach 1:
The patent combines radar and camera data to create a fusion track for stationary target detection. The radar provides distance and velocity information while the camera provides angle and visual confirmation, merging their strengths to overcome the limitations of each individual sensor for stationary target detection.
Solution Approach 2:
The patent introduces a fusion track as an intermediary representation that integrates information from both radar and camera sensors. This fusion track serves as a mediator that combines the complementary data from both sensors to improve measurement precision without directly managing the complexity of multiple sensors.
2Measurement precision
If sensor fusion technology is implemented to improve environment recognition, then the measurement precision improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent extracts only the necessary features from radar and camera data for fusion track generation. Instead of processing all raw data from both sensors, it selectively extracts relevant information such as position, velocity, and visual characteristics to reduce computational complexity while maintaining precision.
Solution Approach 2:
The patent segments the sensor fusion process into distinct stages: generating fusion track from radar and camera data, determining stationary state, collecting radar points, and obtaining center point or generating contour. This segmentation allows each stage to be optimized independently, reducing overall computational complexity.
3Reliability
If fusion track is used for stationary target detection, then the reliability improves, but the loss of time for processing increases
Solution Approach 1:
The patent performs preliminary actions by continuously maintaining a fusion track that integrates radar and camera data in advance. This pre-integrated fusion track is readily available when stationary target detection is needed, eliminating the need for time-consuming real-time fusion processing at the moment of detection.
Solution Approach 2:
The patent maintains continuous fusion track generation and updating from both radar and camera sensors. This continuous operation ensures that the fusion track is always current and ready for immediate use in stationary target detection, eliminating processing delays while maintaining high reliability.
Data Source
AI summary
Provided are a method and device for obtaining a position of a stationary target. The method of obtaining a position of a stationary target may include generating a fusion track based on data collected from a radar and data collected from a camera, determining whether the fusion track is in a stationary state, in response to the fusion track being in the stationary state, collecting radar points associated with the fusion track, and obtaining a center point of the stationary target based on the collected radar points.