Radar-Camera Data Fusion for Resolving 3D Distance Ambiguity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current technologies face challenges in providing an accurate environmental model for autonomous driving and advanced driver-assistance systems, as visual information lacks distance accuracy while radar information lacks object appearance details.

Innovation Solution

A method and system that fuse radar and visual information to generate an accurate environmental model by determining hybrid-detection-based objects using a combination of radar and visual data, solving distance ambiguities and pairing radar and visual objects to provide precise location information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If visual information is used for object detection, then object appearance information is provided, but distance information accuracy deteriorates

Engineering Contradiction:
Improveobject appearance informationVSAvoiddistance information accuracy
Core Design Contradiction:
Loss of informationVSMeasurement precision

Solution Approach 1:

The patent merges visual information from cameras with radar information to create a hybrid detection system. The visual processing unit processes camera data to identify object appearance, while the radar processing unit provides accurate distance measurements. These two information streams are combined in the object detection unit to produce both appearance and precise distance information simultaneously, resolving the contradiction between obtaining appearance details and maintaining distance accuracy.

Inventive Principle:
Principle #5Merging (Combining)

2Measurement precision

If radar information is used for object detection, then distance information accuracy is provided, but object appearance information deteriorates

Engineering Contradiction:
Improvedistance information accuracyVSAvoidobject appearance information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The system combines radar data with visual data from cameras. The radar processing unit extracts accurate distance and velocity information, while the visual processing unit supplements this with object appearance characteristics from camera images. The object detection unit integrates these complementary data sources, allowing the system to achieve both precise distance measurement and object identification simultaneously.

Inventive Principle:
Principle #5Merging (Combining)

3Ease of operation

If separate radar and visual detection systems are used, then each sensor type can operate independently, but environmental model accuracy deteriorates

Engineering Contradiction:
Improveindependent sensor operationVSAvoidenvironmental model accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent implements a fusion architecture where radar and visual detection systems operate independently in their respective processing units, maintaining operational simplicity. However, the object detection unit merges the outputs of both systems to create a unified environmental model. This hybrid approach preserves the independence of individual sensor operations while achieving superior environmental model accuracy through combined data analysis.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS20230228863A1Data fusion for environmental model generation
Publication Date: 2023.07.20 AUTOBRAINS TECH LTD
  • US20230228863A1 patent drawing
  • US20230228863A1 patent drawing
  • US20230228863A1 patent drawing

AI summary

A method for fusion of radar and visual information, the method comprises: obtaining visual information and radar information about a three dimensional (3D) space located within a field of view of a camera that acquired the visual information and within a field of view of a radar that acquired the radar information; finding, based on the visual information, estimated visual-detection-based (VDB) objects and estimated VDB locations of the estimated VDB objects within the 3D space; wherein the estimated VDB locations exhibit a distance ambiguity; determining hybrid-detection-based (HDB) objects and HDB locations of the HDB objects, based on (i) the radar information, (ii) the estimated VDB objects, and (iii) the estimated VDB locations of the VDB objects.