Multi-Directional Phase Difference Detection for Edge Area Accuracy

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing techniques, such as stereo matching, struggle to accurately calculate phase differences in images containing edge areas, leading to incorrect phase difference distributions and subsequent incorrect distance information.

Innovation Solution

An image processing device and method that detects phase differences between parallax images by performing correlation value calculations along multiple directions, generating phase difference distributions by combining results from these directions, thereby reducing the influence of edge areas and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If single-direction phase difference detection is used, then processing simplicity is maintained, but measurement precision deteriorates in edge areas

Engineering Contradiction:
Improvephase difference detection accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent transitions from single-direction phase difference detection to multi-directional detection by adding directional dimension. Specifically, it performs phase difference detection in both horizontal and vertical directions, allowing the system to overcome the limitations of edge areas that cause failure in single-direction detection. This dimensional expansion enables accurate phase difference measurement even in regions where traditional single-direction methods fail.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If multi-directional phase difference detection is performed, then measurement precision improves, but processing time increases

Engineering Contradiction:
Improvephase difference distribution accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by performing edge detection before phase difference detection. The edge detection results are used to identify edge areas that would cause problems in subsequent phase difference detection. This preliminary identification allows the system to selectively apply multi-directional detection only where needed, rather than uniformly across the entire image, thereby reducing overall processing time while maintaining accuracy in critical regions.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If conventional stereo matching is used, then processing simplicity is maintained, but reliability deteriorates due to edge area interference

Engineering Contradiction:
Improvephase difference distribution correctnessVSAvoiddetection method complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces edge detection as an intermediary step between image acquisition and phase difference detection. This intermediary process identifies problematic edge areas that would otherwise cause unreliable phase difference measurements. By using edge detection results to guide the phase difference detection process, the system achieves more reliable results in edge areas without requiring complete redesign of the detection methodology.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS10015472B2Image processing using distance information
Publication Date: 2018.07.03 SONY GROUP CORP
  • US10015472B2 patent drawing
  • US10015472B2 patent drawing
  • US10015472B2 patent drawing

AI summary

An image processing device includes a phase difference detection portion configured to detect phase difference between parallax images by performing correlation value calculation with the parallax images, and generate phase difference distribution in an image. The phase difference detection portion performs the phase difference detection individually along two or more directions different from each other, and generates the phase difference distribution by utilizing results of the phase difference detection regarding the two or more directions.