3D Shape Reconstruction Using Multi-View Distance Image Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for generating three-dimensional shape data, such as the visual hull method, often result in inaccuracies when dealing with objects having curved or concave surfaces, leading to significant errors in the approximation of these shapes.

Innovation Solution

An image processing apparatus and method that utilizes multiple cameras to capture images from different viewpoints, generates approximate shape data using the visual hull method, corrects distance images based on surface three-dimensional information, and applies threshold values to refine the shape data, ensuring high accuracy even for objects with complex shapes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the visual hull method is used to generate three-dimensional shape data, then the generation speed and stability are improved, but the accuracy deteriorates for curved or concave surfaces

Engineering Contradiction:
Improvegeneration speedVSAvoidshape accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the three-dimensional space into multiple local coordinate systems, each responsible for a specific region. By dividing the overall shape restoration task into regional sub-tasks, the system can apply specialized processing to different areas, improving accuracy for complex surfaces while maintaining overall efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality by using different coordinate systems and processing strategies for different regions. Each local coordinate system is optimized for its specific region, allowing the system to handle curved and concave surfaces with higher precision in those areas while maintaining general efficiency.

Inventive Principle:
Principle #3Local quality

2Stability of the object's composition

If the visual hull method is used to generate three-dimensional shape data, then the processing stability is improved, but the accuracy for concave portions deteriorates

Engineering Contradiction:
Improveprocessing stabilityVSAvoidconcave portion accuracy
Core Design Contradiction:
Stability of the object's compositionVSMeasurement precision

Solution Approach 1:

The patent divides the processing space into multiple local coordinate systems, each stable for its region. This segmentation allows the system to maintain the stability of the visual hull method overall while improving accuracy in specific problematic regions through localized processing.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

By applying local quality principles, the patent uses specialized coordinate systems and processing methods for regions with concave portions. This allows the system to maintain processing stability globally while achieving high accuracy locally for difficult geometric features.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If multiple coordinate systems are used to improve accuracy, then the measurement precision is improved, but the device complexity increases

Engineering Contradiction:
Improveshape data accuracyVSAvoidcoordinate system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent manages complexity by segmenting the coordinate system into manageable local units. Each local coordinate system handles a specific region, making the overall complex system decomposable and easier to implement and maintain while achieving high precision.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4336448B1Image processing apparatus, image processing method, program, and storage medium
Publication Date: 2026.04.22 CANON KK
  • EP4336448B1 patent drawingFigure 1
  • EP4336448B1 patent drawingFigure 2
  • EP4336448B1 patent drawingFigure 3

AI summary

An object is to obtain three-dimensional shape data with high accuracy from three-dimensional shape data representing an approximate shape of an object. Three-dimensional shape data of an object existing in an image capturing space is obtained (302). Further, a plurality of distance images each representing a distance to the object and corresponding to a plurality of viewpoints is obtained (305). Then, correction to evaluate the three-dimensional shape data based on the plurality of distance images and based on results of the evaluation, delete a unit element estimated not to represent a shape of the object among unit elements configuring the three-dimensional shape data is performed (307).