Stereo Image Parallax Correction Using Mark Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for determining the depth position of a photographed scene from stereo images are inefficient, leading to inaccuracies and long processing times, especially when camera parameters change, and fail to correct depth positions accurately when focal lengths differ between cameras.

Innovation Solution

An information processing apparatus and method that acquires and corrects stereo images using a mark detection unit, parallax determination unit, and correction unit, which calculates and applies correction values to adjust image sizes and positions, enabling accurate depth position determination with reduced processing time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If stereo rectification is performed using camera parameters to facilitate the search for corresponding positions, then the search efficiency is improved, but when camera parameters change, the depth position accuracy deteriorates

Engineering Contradiction:
Improvesearch efficiencyVSAvoiddepth position accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary stereo rectification using initial camera parameters to establish a baseline coordinate system and epipolar geometry. This preliminary action enables efficient searching while the system later compensates for parameter changes through correction mechanisms that adjust for drift from the initial rectification state.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback by continuously monitoring camera parameter changes and correcting the rectification results accordingly. When camera parameters change, the system calculates correction values based on the difference between current and initial parameters, then applies these corrections to maintain depth position accuracy despite parameter variations.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If the photographed image is shifted in the height direction by one pixel at a time to reduce positional displacement, then the depth position accuracy is improved, but the processing time increases significantly

Engineering Contradiction:
Improvedepth position accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Instead of iteratively shifting images by one pixel at a time, the patent calculates the required height direction shift amount directly using camera parameter changes. This allows the system to jump to the correct position in a single step rather than incrementally searching, dramatically reducing processing time while maintaining accuracy.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical iterative shifting approach with a computational solution that uses camera parameter differential calculations. Instead of physically or digitally moving images step-by-step, the system computes the exact correction needed based on parameter changes, substituting brute-force iteration with mathematical derivation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If a search is performed on the whole of the other photographed image for each pixel to determine corresponding positions, then the depth position accuracy is improved, but the processing complexity and time increase

Engineering Contradiction:
Improvedepth position accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts and utilizes the epipolar constraint relationship to reduce the search space from the entire image to a single epipolar line. By taking out and applying this geometric constraint, the system determines corresponding positions without exhaustive full-image searching, significantly reducing complexity while maintaining accuracy through the epipolar geometry inherent in stereo vision.

Inventive Principle:
Principle #2Taking out (Extraction)

4Speed

If stereo rectification is performed based on initial camera parameters, then the processing speed is improved, but when focal lengths differ between cameras, the depth position accuracy deteriorates

Engineering Contradiction:
Improveprocessing speedVSAvoiddepth position accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent performs preliminary stereo rectification using initial camera parameters to establish a baseline coordinate system and enable fast processing. This preliminary rectification provides a good approximation that works well when parameters are stable, while the system later applies corrections to handle focal length differences and other parameter variations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback by detecting camera parameter changes (particularly focal length variations) and calculating correction values based on these changes. The correction mechanism adjusts the rectification results to compensate for parameter drift, ensuring accuracy is maintained even when focal lengths differ between cameras or change over time.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS20230260159A1Information processing apparatus, information processing method, and non-transitory computer readable medium
Publication Date: 2023.08.17 CANON KK
  • US20230260159A1 patent drawing
  • US20230260159A1 patent drawing
  • US20230260159A1 patent drawing

AI summary

An information processing apparatus acquires two images; detects a mark from each of the images; acquires a correction value for at least one of the images on a basis of a positional relationship in a height direction between the marks in the images; determines a parallax between the images on a basis of the images and the correction value; and corrects the parallax or a depth position corresponding to the parallax on a basis of a positional relationship between the marks in the images and the correction value, wherein the correction value includes at least one of a correction value for enlarging or reducing at least one of the images and a correction value for moving at least one of the images in the height direction.