Radar Object Detection Grid Clustering with Resolution Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Clustering operations using DBSCAN become challenging when radars with different spatial resolutions detect the same object from overlapping measurement areas, making it difficult to set proper parameters for clustering operations, especially when radars are arranged to face each other at intersections.
Innovation Solution
An object detection device and method that calculate likelihoods for radar cells based on distances and perform clustering operations using grid values, allowing for proper clustering even when radars have different spatial resolutions and overlapping measurement areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple radars with different spatial resolutions are used to detect objects from overlapping measurement areas, then the coverage and detection capability are improved, but it becomes difficult to set proper parameters for clustering operations
Solution Approach 1:
The patent introduces an intermediary transformation process that converts measurement data from radars with different spatial resolutions into a unified coordinate system with standardized cell dimensions. This intermediary step acts as a mediator that harmonizes the different resolution formats, allowing clustering operations to proceed with consistent parameters regardless of the original radar resolution variations.
Solution Approach 2:
The patent applies parameter changes by transforming the spatial resolution parameters of multiple radars into a common reference frame. By changing the coordinate system parameters and standardizing cell sizes through mathematical transformation, the system enables clustering operations to use uniform parameters (ε and minPts) across all radar data, resolving the parameter setting difficulty.
2Adaptability or versatility
If radars are arranged to face each other at intersections to detect objects from different directions, then the detection coverage is improved, but the radars cannot share common measurement points effectively
Solution Approach 1:
The patent resolves the measurement point sharing issue by introducing a dimensional transformation approach. It converts measurement data from different radar coordinate systems into a unified three-dimensional spatial coordinate system, allowing measurement points from radars facing different directions to be accurately correlated. This dimensional change enables effective data fusion even when radars are positioned at intersections facing each other.
3Productivity
If DBSCAN clustering algorithm is used on raw radar measurement data with different spatial resolutions, then the clustering operation becomes computationally intensive, but proper clustering cannot be achieved without parameter adjustment
Solution Approach 1:
The patent applies preliminary action by performing coordinate transformation and spatial resolution standardization before the clustering operation. This pre-processing step converts all radar measurement data into a unified coordinate system with consistent cell dimensions, so that the subsequent DBSCAN clustering can be executed with fixed parameters without requiring iterative adjustment, thereby improving both efficiency and accuracy.
Data Source
AI summary
A processor in an object detection device is configured to: acquire settings data for one or more radar grids, each radar grid being set for a measurement area of a corresponding radar and consisting of radar cells; acquire settings data for a processing grid for clustering operations, the processing grid being set for the one or more measurement areas and consisting of processing cells; calculate, for each processing cell, one or more likelihoods associated with measurements of one or more related radar cells based on distances between the processing cell and the one or more related radar cells; calculate a grid value of each of the processing cells in the processing grid based on the one or more likelihoods; and perform a clustering operation on each processing cell based on distances between and/or grid values of the processing cell and one or more different processing cells.