Radar Object Detection Using Concatenated Point and Map Data

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for object detection using radar data in automotive perception systems are limited in predicting the location and motion of objects, and there is a need for enhanced detection performance.

Innovation Solution

A computer-implemented method that combines radar detection data and map data into suitable representations for concatenation, using a grid system to process radar sensor data, removing ego-location effects, and converting data into point and map representations to enhance object detection through neural networks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional radar object detection methods are used, then the system is simple to implement, but the detection performance and accuracy are limited

Engineering Contradiction:
Improvedetection performanceVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines radar detection data with map data into a unified representation, merging multiple data sources to enhance detection performance. The concatenated representation integrates point cloud information from radar with spatial context from maps, improving object detection accuracy beyond what conventional single-source methods achieve.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces a new dimensional representation by concatenating radar detection data with map data along a feature dimension. This creates an enriched data structure that adds spatial and contextual dimensions to traditional radar point clouds, enabling more sophisticated object detection without fundamentally changing the radar hardware.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Loss of information

If radar detection data and map data are combined into concatenated representations, then information content increases for better detection, but data processing complexity increases

Engineering Contradiction:
Improveinformation contentVSAvoiddata processing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent performs preliminary processing of both radar detection data and map data before concatenation, converting them into compatible representations. By pre-processing and standardizing the data formats in advance, the system reduces the computational burden during real-time detection while preserving all essential information from both sources.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3767332B1Methods and systems for radar object detection
Publication Date: 2023.12.13 APTIV TECHNOLOGIES LTD
  • EP3767332B1 patent drawingFigure 1
  • EP3767332B1 patent drawingFigure 2
  • EP3767332B1 patent drawingFigure 3

AI summary

A computer implemented method for object detection comprises the following steps: determining a grid, the grid comprising a plurality of grid cells; determining, for a plurality of time steps, for each grid cell, a plurality of respective radar detection data, each radar detection data indicating a plurality of radar properties; determining, for each time step, a respective radar map indicating a pre-determined radar map property in each grid cell; converting the respective radar detection data of the plurality of grid cells for the plurality of time steps to a point representation of pre-determined first dimensions; converting the radar maps for the plurality of time steps to a map representation of pre-determined second dimensions, wherein the pre-determined first dimensions and the pre-determined second dimensions are at least partially identical; concatenating the point representation and the map representation to obtain concatenated data; and carrying out object detection based on the concatenated data.