Stereo Depth Mapping with Restricted Search Ranges

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing three-dimensional measurement methods based on stereo matching are prone to inaccuracies due to variations in reflection characteristics and environmental changes, leading to reduced measurement accuracy and reliability.

Innovation Solution

An image processing device that generates a depth map using stereo matching with a restricted search range based on disparities estimated from multiple image pairs taken from different viewpoints or under different conditions, combining these disparities to enhance accuracy and robustness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If stereo matching is performed over a wide search range to ensure comprehensive correspondence search, then measurement reliability is improved, but measurement time increases significantly

Engineering Contradiction:
Improvemeasurement reliabilityVSAvoidmeasurement time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by performing a first disparity estimation using a coarse method (such as block matching or gradient-based methods) before conducting the final precise stereo matching. This preliminary disparity map narrows down the search range for subsequent detailed correspondence search, ensuring that the comprehensive search is confined to a restricted area around the preliminary estimate, thus maintaining reliability while significantly reducing measurement time.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If multiple image pairs are processed to reduce measurement variations, then measurement accuracy is improved, but processing complexity increases

Engineering Contradiction:
Improvemeasurement accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple disparity maps obtained from different image pairs through a composite disparity map generation process. The disparity estimation results from multiple stereo image pairs are combined using techniques such as averaging or weighted averaging, which reduces measurement variations and improves accuracy. This merging process consolidates multiple measurements into a single robust result, managing processing complexity through systematic integration.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If the search range is restricted based on preliminary disparity estimation to reduce processing time, then measurement speed is improved, but risk of missing correct correspondence increases

Engineering Contradiction:
Improvemeasurement speedVSAvoidcorrespondence search reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies local quality by adjusting the search range dynamically based on local disparity characteristics. Instead of using a uniform fixed search range, the search area is adapted locally around each pixel's preliminary disparity estimate, with the range potentially varying by location. This ensures that the correct correspondence is not missed while restricting the search to necessary areas, thus improving measurement speed without sacrificing reliability.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3832601B1Image processing device and three-dimensional measuring system
Publication Date: 2025.09.03 OMRON CORP
  • EP3832601B1 patent drawingFigure 1
  • EP3832601B1 patent drawingFigure 2
  • EP3832601B1 patent drawingFigure 3

AI summary

An image processing device includes a setting unit configured to estimate a disparity between a first image and a second image based on a method different from stereo matching and set, based on the disparity thus estimated, a search range for a corresponding point for the stereo matching, a disparity map generating unit configured to search only the search range thus set for a corresponding point of each pixel between the first image and the second image and generate a disparity map, and a disparity map compositing unit configured to composite a plurality of the disparity maps each generated from a corresponding one of a plurality of image pairs.