Depth-Guided Image Processing for Accurate 3D Model Regions

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for generating three-dimensional models from captured images fail to accurately confirm which parts of an object are within a predetermined range, such as an in-focus range, leading to potential blurring and inaccurate separation of objects outside this range.

Innovation Solution

An image processing system that utilizes depth information to distinguish between objects within a predetermined distance range and those outside it, displaying them differently and specifying partial regions used for three-dimensional model generation, using a combination of foreground separation, depth information, and shape estimation techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a captured image is used for three-dimensional model generation without depth information filtering, then the object can be captured even outside the in-focus range, but the object cannot be accurately separated and the model generation precision deteriorates

Engineering Contradiction:

Improveability to capture objects outside in-focus rangeVSAvoidobject separation accuracy

Core Design Contradiction:

Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The system performs preliminary depth information filtering before object separation. By determining whether each captured image contains an object within the predetermined depth range (in-focus range) in advance, the system pre-selects suitable images for three-dimensional model generation, preventing inaccurate separation of out-of-focus objects from affecting the model.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Depth information acts as an intermediary between the captured image and the three-dimensional model generation process. The depth information filter serves as a mediator that determines whether an object should be included in the model generation, separating the tasks of image capture and model generation by inserting a depth verification step in between.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If depth information filtering is applied to select only in-focus objects for model generation, then the model generation precision is improved, but the ability to confirm which parts of captured images are within the predetermined range is lost

Engineering Contradiction:

Improvethree-dimensional model generation precisionVSAvoidvisibility of in-focus range information

Core Design Contradiction:

Manufacturing precisionVSLoss of information

Solution Approach 1:

The system visually distinguishes between different depth ranges by applying different display characteristics to captured images. Images containing objects within the predetermined depth range are displayed with one set of characteristics (e.g., normal brightness), while images with objects outside the depth range are displayed with different characteristics (e.g., adjusted brightness or filtering), making the depth information visible to the user.

Inventive Principle:

Principle #32Color changes

Solution Approach 2:

The system provides feedback to the user about which captured images contain objects within the predetermined depth range. By displaying depth range information or visually indicating the in-focus regions, the system allows users to understand and verify which parts of the captured images are suitable for three-dimensional model generation.

Inventive Principle:

Principle #23Feedback

3Device complexity

If captured images are processed without depth information, then the processing flow is simple, but unnecessary parameter adjustments are required and productivity decreases

Engineering Contradiction:

Improveimage processing flow complexityVSAvoidthree-dimensional model generation efficiency

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The system performs depth information-based filtering as a preliminary step before the main three-dimensional model generation process. By pre-determining which captured images contain objects within the predetermined depth range, the system eliminates the need for subsequent parameter adjustments and reduces unnecessary processing, thereby improving productivity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system extracts and removes captured images that do not meet the depth range requirement from the processing pipeline. By taking out unnecessary images (those with objects outside the in-focus range) before they enter the model generation process, the system reduces the workload and eliminates unnecessary parameter adjustments, improving overall efficiency.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20260045034A1Image processing system, control method, and storage medium

Publication Date: 2026.02.12 CANON KK

AI summary

An image processing system acquires an image based on image capturing by an image capturing device and depth information indicating a predetermined distance range from the image capturing device and displays, in different display manners, a first object present in the predetermined distance range and a second object not present in the predetermined distance range among objects in the image.