Bone-Guided 3D Model Generation Under Multi-Player Occlusion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Occlusion in 3D model generation due to limitations in camera arrangement and number of players in volumetric image capture environments, such as sports games, leads to inaccurate 3D model generation.

Innovation Solution

An image processing device and method that estimates bone information from captured images and generates a 3D model based on this information, even when occlusion occurs, using a bone information estimation unit and a 3D model estimation unit to handle occluded areas by deforming rigged 3D models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If volumetric image capture is performed with limited camera arrangement and number of cameras, then device complexity is reduced, but measurement precision of 3D model generation deteriorates due to occlusion

Engineering Contradiction:
Improvecamera arrangement complexityVSAvoid3D model generation accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system performs preliminary actions by capturing images at multiple time points before the occlusion occurs, storing these images in a buffer. When occlusion is detected at a predetermined time, the system retrieves previously captured non-occluded images from the buffer to generate the 3D model, ensuring accurate measurement despite current occlusion conditions.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates a copy of the 3D model generation process by using previously captured images as substitutes for the current occluded images. This copying approach allows the system to maintain measurement precision by referring to accurate historical data rather than relying on compromised current data.

Inventive Principle:
Principle #26Copying

2Adaptability or versatility

If the number of players to be generated is increased, then adaptability of the system improves, but measurement precision deteriorates due to occlusion between players

Engineering Contradiction:
Improvenumber of players handledVSAvoid3D model accuracy per player
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system captures images of all players at multiple time points before occlusion occurs, storing these images in association with each player's identification information. When occlusion is detected during processing, the system retrieves previously captured non-occluded images from the buffer to maintain accurate 3D model generation for each player, regardless of the number of players being processed simultaneously.

Inventive Principle:
Principle #10Preliminary action

3Productivity

If 3D model generation is performed in real-time, then productivity is improved, but measurement precision deteriorates when occlusion occurs

Engineering Contradiction:
Improvereal-time processing speedVSAvoid3D model accuracy under occlusion
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary image capture at multiple time points and stores these images in a buffer before occlusion occurs. During real-time processing, when occlusion is detected, the system quickly retrieves pre-captured non-occluded images from the buffer, maintaining both high productivity through real-time processing and high measurement precision by using accurate historical image data.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS20250285378A1Image processing device, image processing method, and program
Publication Date: 2025.09.11 SONY GROUP CORP
  • US20250285378A1 patent drawing
  • US20250285378A1 patent drawing
  • US20250285378A1 patent drawing

AI summary

The present disclosure relates to an image processing device, an image processing method, and a program that are capable of generating a 3D model with high accuracy even when occlusion has occurred.The image processing device includes: a bone information estimation unit that estimates, based on captured images at a predetermined time captured by a plurality of imaging devices, bone information of a person appearing in the captured images; and a 3D model estimation unit that, when occlusion has occurred on the person appearing in the captured images at the predetermined time, estimates a 3D model of the person based on the bone information at the predetermined time. The technology of the present disclosure can be applied to, for example, an image processing device that distributes free viewpoint video.