Multi-Radar Image Synthesis for Autonomous Vehicle Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current radar systems for autonomous vehicles face challenges in generating high-resolution images of surroundings in real-time, especially when detecting objects at high speeds or in varying environments, due to limitations in beamforming and signal processing capabilities.

Innovation Solution

The method involves synchronizing multiple radars to perform beamforming on the same point simultaneously, synthesizing overlapping images to achieve a 360-degree field of view, and adjusting gains and beamforming ranges based on vehicle modes (e.g., highway or city driving) to enhance object detection and recognition.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple radars are used to improve detection range and resolution, then measurement precision and detection range are improved, but device complexity increases

Engineering Contradiction:
Improveimage resolutionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

Multiple radar captured images are merged into a single synthesized image through coordinate transformation and overlapping region integration. This combining approach achieves high-resolution imaging with extended detection range while managing system complexity through unified processing architecture

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The radar image processing apparatus performs multiple functions including coordinate transformation, image synthesis, object detection, and high-resolution image generation using a single integrated system. This multi-functionality reduces the need for separate specialized components

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Measurement precision

If radars perform beamforming on the same point simultaneously, then measurement precision and image quality are improved, but loss of time occurs due to synchronization requirements

Engineering Contradiction:
Improvebeamforming accuracyVSAvoidsynchronization time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Transmission times for transmitting transmission signals from the radars toward a target are synchronized in advance, and reception times for receiving reflection signals are synchronized beforehand. This preliminary synchronization ensures precise simultaneous beamforming without real-time delays

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses synchronized transmission and reception timing with feedback mechanisms to coordinate multiple radars. The synchronization information is used to adjust beamforming operations to achieve precise spatial-temporal alignment

Inventive Principle:
Principle #23Feedback

3Measurement precision

If the beamforming range is extended to cover all directions, then detection range is improved, but use of energy increases

Engineering Contradiction:
Improvedetection rangeVSAvoidenergy consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The beamforming range is dynamically adjusted based on vehicle mode (highway or city driving). During highway driving, the beamforming range is limited to the front side to reduce energy consumption, while during city driving it covers all directions for comprehensive detection

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters including beamforming range and gain settings based on driving conditions. Transmission signals are coded differently and gains are adjusted to optimize energy usage while maintaining detection performance

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3457173B1Radar image processing method, apparatus, and system
Publication Date: 2023.09.06 SAMSUNG ELECTRONICS CO LTD
  • EP3457173B1 patent drawingFigure 1
  • EP3457173B1 patent drawingFigure 2A
  • EP3457173B1 patent drawingFigure 2B

AI summary

A radar image processing method includes acquiring captured images from radars synchronized to perform beamforming on a same point at a same time, synthesizing the captured images based on at least one overlapping area of the captured images, and generating a high-resolution image based on the synthesized images.