Stereo Camera Rolling Correction for Road Surface Shape Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional systems that estimate road surface shape using distance images struggle with accuracy when rolling occurs, especially on surfaces with changing gradients, as they rely on assumptions of a flat road surface, leading to inaccurate road surface shape estimation.

Innovation Solution

The system employs a road surface estimating unit that generates a V map and corrects for the influence of rolling by calculating and applying rolling correction amounts to the height table, ensuring accurate road surface shape estimation even on surfaces with gradients.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a correction process based on the assumption of a flat road surface is used, then the processing complexity is reduced, but the measurement precision of road surface shape deteriorates when rolling occurs on surfaces with changing gradients

Engineering Contradiction:
Improvecorrection process complexityVSAvoidroad surface shape estimation accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent changes the parameter used for correction from a flat road surface assumption to a curved road surface model. By introducing curvature parameters (κx, κy) to describe the road surface shape and using them in the correction process, the system can accurately correct rolling effects even on surfaces with changing gradients, thereby improving measurement precision without excessive complexity increase

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent makes the correction process dynamic by adapting it to the actual road surface conditions. Instead of using a static flat surface assumption, the system dynamically adjusts the correction model based on detected road surface curvature and rolling angles, allowing accurate estimation under varying conditions while maintaining reasonable processing complexity

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If rolling correction is applied to eliminate the influence of rolling, then the measurement precision improves, but the device complexity increases due to additional correction processes

Engineering Contradiction:
Improveroad surface shape estimation accuracyVSAvoidcorrection process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary estimation of road surface curvature and rolling angles before the main correction process. By pre-calculating these parameters from the distance image data, the system prepares the necessary correction values in advance, which streamlines the subsequent correction process and reduces overall computational complexity while maintaining high measurement precision

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces intermediate parameters (road surface curvature κx, κy and rolling angles θx, θy) that act as mediators between the raw distance image data and the final corrected road surface shape estimation. These intermediary parameters decouple the complex correction process into manageable steps, improving precision without proportionally increasing device complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3549055B1Information processing device, imaging device, apparatus control system, information processing method, and computer program product
Publication Date: 2023.06.28 RICOH CO LTD
  • EP3549055B1 patent drawingFigure 1~2
  • EP3549055B1 patent drawingFigure 3~4
  • EP3549055B1 patent drawingFigure 5

AI summary

An information processing device includes a disparity information input unit 151 configured to acquire disparity image data indicating a disparity image including a disparity; a V map generating unit 152 configured to generate a V map indicating a frequency of the disparity for each of coordinates of the disparity image by voting the disparity at a position corresponding to the coordinate of each of pixels of the disparity image based on the disparity image data; a shape estimating unit 156 configured to estimate a road surface shape based on the V map; a height table generating unit 157 configured to generate a height table indicating a relationship between the disparity and a height of a road surface based on the estimated road surface shape; and a correcting unit 161 configured to correct, based on rolling information on rolling of a stereo camera that acquires the disparity, information used to generate information indicating the road surface shape so as to cancel out an influence of the rolling.