Vehicle Radar Monitoring Using Occupancy Maps for Stationary Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle radar systems struggle with accurately detecting and distinguishing stationary objects from moving objects, leading to false detections and missed detections due to varying signal characteristics and overlapping detection areas, which can compromise the safety and reliability of driver assistance systems.
Innovation Solution
A vehicle monitoring apparatus and method using a sensor unit with detection sensors and a controller to extract stationary objects, map them to a grid map, calculate occupancy probability parameters, and correct shaded grids based on vehicle speed, applying clustering algorithms to recognize continuous structures and free spaces, thereby enhancing the accuracy of environmental monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If general radar detection is used to detect outside objects, then the detection area is covered, but stationary objects cannot be accurately distinguished from moving objects leading to false detections
Solution Approach 1:
The patent segments the detection process by separating stationary object detection from moving object detection. It uses multiple radar signals with different characteristics (different transmission directions, frequencies, or waveforms) to detect the same area, allowing the system to distinguish stationary objects by comparing detection results across different signals. This segmentation enables accurate identification of stationary objects while reducing false detections of moving objects.
Solution Approach 2:
The patent changes detection parameters by using multiple radar signals with varying characteristics (transmission direction, frequency, waveform) to detect the same spatial area. By analyzing occupancy probability parameters across these different signals and time frames, the system can differentiate between stationary objects (which maintain consistent occupancy across frames) and moving objects (which show changing occupancy patterns), thereby improving detection accuracy and reducing false alarms.
2Measurement precision
If multiple radar signals are used to detect stationary objects, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The patent creates a grid map that serves multiple functions: it tracks stationary objects, monitors moving objects, and provides a unified framework for integrating multiple radar signal detections. The grid map structure allows different radar signals to contribute to a common representation of the environment, enabling the system to handle multiple detection tasks through a single unified data structure and processing approach, thereby managing complexity while maintaining multi-signal detection capabilities.
Solution Approach 2:
The patent uses occupancy probability parameters as simplified representations (copies) of complex radar detection data. Instead of processing raw radar signals directly, the system converts detection results into occupancy probability values for each grid cell, which can be easily compared and integrated across multiple signals and time frames. This copying approach simplifies the processing of multiple radar signals while preserving the essential information needed for stationary object detection.
3Measurement precision
If occupancy probability calculation is performed across multiple frames, then stationary object identification improves, but processing time increases
Solution Approach 1:
The patent applies partial action by focusing occupancy probability calculations only on grid cells that show consistent occupancy patterns across multiple frames. Instead of processing the entire grid map for every frame, the system identifies and processes only the relevant regions where stationary objects are likely to be located, reducing the overall computational burden while maintaining accurate stationary object identification.
Solution Approach 2:
The patent performs preliminary filtering by identifying consistently occupied grid cells before conducting detailed occupancy probability calculations. By pre-identifying potential stationary object locations through simple occupancy tracking across frames, the system reduces the scope of subsequent complex probability calculations to only those relevant areas, thereby reducing processing time while maintaining identification accuracy.
Data Source
AI summary
An apparatus for monitoring the surrounding environment of a vehicle may include a sensor unit including a plurality of detection sensors configured to detect an object outside the vehicle according to frames with a time period; and a controller configured to extract a stationary object among outside objects detected through the sensor unit based on behavior information of the vehicle, map the extracted stationary object to a grid map, add occupancy information to each of grids constituting the grid map, in response to the stationary object being mapped to the grid map, calculate an occupancy probability parameter indicating the probability that the stationary object will be located at each of the grids, from the occupancy information added to the grids within the grid map in a plurality of frames, and monitor the surrounding environment of the vehicle on the basis of the occupancy probability parameter.