Virtual Viewpoint Video Object Persistence via RFID Positional Data

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In virtual viewpoint video image generation, objects that do not move for a predetermined time are incorrectly identified as background and thus disappear from the generated three-dimensional shape data, leading to their sudden disappearance in the virtual viewpoint video image.

Innovation Solution

An image processing apparatus acquires positional information from RF tags or similar notification units attached to equipment objects using RFID antennas, transforming this information into three-dimensional shape data to maintain the representation of these objects in the virtual viewpoint video image space, even if they are stationary.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the background difference method is used to update the background based on newly captured video images, then the background can be updated to reflect changes in illumination conditions, but stationary objects that do not move for a predetermined time are incorrectly identified as background and disappear from the virtual viewpoint video image

Engineering Contradiction:

Improvebackground update capabilityVSAvoidobject detection accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent introduces positional information notification units (such as RFID tags, barcodes, or QR codes) as intermediaries attached to stationary objects. These units provide additional identification information that allows the system to distinguish between actual background and stationary foreground objects, preventing the latter from being incorrectly identified as background during the background difference method processing

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent attaches positional information notification units to stationary objects in advance before video capture begins. This preliminary action ensures that when background difference method is applied, the system can immediately recognize these objects through their notification unit information and prevent them from being misidentified as background, thus maintaining their presence in the virtual viewpoint video image throughout the process

Inventive Principle:

Principle #10Preliminary action

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Prevents the sudden disappearance of objects from the virtual viewpoint video image by ensuring that stationary equipment objects are correctly identified and represented in the three-dimensional model, allowing for a more stable and natural virtual viewpoint video image generation.

Implementation Method 1

positional information on equipment objects from RF tags attached to the equipment objects

Methodology Applied

Scientific EffectRFID (Radio Frequency Identification): Electromagnetic Induction

Data Source

PatentUS11665332B2Information processing apparatus, control method thereof and storage medium

Publication Date: 2023.05.30 CANON KK

AI summary

To prevent an object that should exist in a virtual viewpoint video image from disappearing suddenly. Positional information on a specific object is acquired from a positional information notification unit included in the specific object. Then, based on the positional information, three-dimensional shape data on the specific object is output in association with information specifying a position of the specific object in a three-dimensional space.