LiDAR-Camera Signal Format Conversion for Robust Distance Measurement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing distance measurement technologies face challenges in combining signals from different formats, such as LiDAR and stereo cameras, leading to inconsistent performance across various scenes and environments, requiring human intervention to set logic for signal combination.

Innovation Solution

A distance measuring apparatus that uses machine learning to combine signals from different formats by converting the format of LiDAR signals to match that of camera signals, allowing a neural network to learn the correspondence between input data and output distance information without human-set logic, thereby achieving robust distance measurement across various scenes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If LiDAR and stereo camera signals are combined using traditional fusion methods, then distance measurement precision is improved, but the system requires human intervention to set logic for signal combination, reducing adaptability

Engineering Contradiction:
Improvedistance measurement precisionVSAvoidadaptability to different scenes
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system employs a neural network that automatically learns and determines the optimal logic for combining LiDAR and stereo camera signals without requiring human intervention. The neural network processes the multi-format signals and autonomously generates distance information, enabling the system to adapt to various scenes and environments independently.

Inventive Principle:
Principle #25Self-service

2Ease of manufacture

If signals in different formats are combined without format conversion, then device complexity is reduced, but the neural network cannot effectively learn correspondence between input data and output distance information

Engineering Contradiction:
Improvesystem implementation easeVSAvoidinformation correspondence
Core Design Contradiction:
Ease of manufactureVSLoss of information

Solution Approach 1:

The patent introduces a signal conversion unit as an intermediary component that transforms LiDAR signals into a format compatible with stereo camera data. This conversion enables the neural network to effectively process and learn from the combined signals while preserving the essential distance measurement information from both sources.

Inventive Principle:
Principle #24Intermediary (Mediator)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The apparatus provides highly robust and precise distance measurement by extracting feature values that exceed human-set logic, enabling consistent performance across diverse scenes and environments without requiring human intervention in logic setting.

Implementation Method 1

the distance to an object is calculated based on the length of time it takes from a point in time when a laser beam is emitted to an object until a point in time when a reflected signal returns from the object

Methodology Applied
Scientific EffectTime of flight: Time of Flight

Data Source

PatentEP3764124A1Distance measuring apparatus, method for measuring distance, on-vehicle apparatus, and mobile object
Publication Date: 2021.01.13 RICOH CO LTD
  • EP3764124A1 patent drawingFigure 1
  • EP3764124A1 patent drawingFigure 2
  • EP3764124A1 patent drawingFigure 3

AI summary

A distance measuring apparatus (100) includes a signal conversion unit (22) to match a format of a second signal relating to a distance to an object with a format of a first signal relating to an image; and a distance information output unit (24) to output distance information based on the first signal and the second signal having the format matched with the format of the first signal. The signal conversion unit (22) converts the second signal such that, among length information in directions of a plurality of dimensions held in the first signal and length information in directions of a plurality of dimensions held in the second signal, length information in a direction of at least one of the plurality of dimensions of the first signal represents length information in a direction of at least one of the plurality of dimensions of the second signal.