Virtual Registered Sphere for 3D Image Capture Guidance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for generating datasets of three-dimensional objects require users to have intimate knowledge of feature extraction and processing, making it demanding and less intuitive for users to determine if images are captured to a satisfactory degree from various angles, especially in Augmented Reality applications.

Innovation Solution

A method that generates a virtual registered sphere divided into contiguous patches, each corresponding to a continuous range of image capture angles, providing visual guidance by rendering the sphere in the camera view and assigning colors to patches based on the quality of captured images, indicating whether they are captured to a satisfactory degree.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If known methods are used for dataset generation, then the dataset can be generated with sufficient technical detail, but the user requires intimate knowledge regarding feature extraction and related processing

Engineering Contradiction:
Improvedataset qualityVSAvoiduser knowledge requirement
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent introduces an intermediary visualization system (virtual registered sphere with color-coded patches) that mediates between the complex feature extraction processes and the user. This intermediary translates technical dataset quality metrics into intuitive visual feedback, allowing users to assess capture quality without needing intimate knowledge of feature extraction algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements real-time feedback by continuously updating the visualization of the virtual registered sphere based on the quality of captured images from different angles. The color changes in patches provide immediate feedback to users about whether images have been captured to a satisfactory degree, enabling users to adjust their capture process accordingly without understanding the underlying technical metrics.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If detailed feature extraction processing is performed, then the dataset accuracy is improved, but the complexity of the process increases

Engineering Contradiction:
Improveimage capture quality assessmentVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a virtual copy (virtual registered sphere) that replicates the essential information about image capture quality from different angles. Instead of requiring users to directly interpret complex feature extraction results, the system generates a simplified visual representation that copies the critical quality metrics in an easily understandable format, maintaining measurement precision while reducing processing complexity perception.

Inventive Principle:
Principle #26Copying

3Reliability

If images are captured from many different angles, then the completeness of the dataset is improved, but the time required for capture increases

Engineering Contradiction:
Improvedataset completenessVSAvoidimage capture time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system applies partial action by allowing users to stop capturing images once sufficient coverage is achieved, as indicated by the visualization. The virtual registered sphere shows which angular ranges have been adequately captured (green patches) and which need more work (red/yellow patches), enabling users to perform just enough capturing to achieve dataset completeness without excessive time investment in capturing redundant images.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS9998655B2Visualization for viewing-guidance during dataset-generation
Publication Date: 2018.06.12 QUALCOMM INC
  • US9998655B2 patent drawing
  • US9998655B2 patent drawing
  • US9998655B2 patent drawing

AI summary

Disclosed is a method and apparatus for providing visual guidance to a user capturing images of a three-dimensional object. In one embodiment, the operations implemented may include: generating a virtual registered sphere comprising a plurality of contiguous patches, wherein each of the plurality of patches corresponds to a continuous range of image capture angles; rendering at least a portion of the virtual registered sphere in an image capture camera view; determining whether images of the three-dimensional object have been captured to a predetermined satisfactory degree from a particular range of image capture angles associated with a particular patch; and assigning a color to the particular patch based at least in part on the determination of whether images of the three-dimensional object have been captured to the predetermined satisfactory degree from the particular range of image capture angles associated with the particular patch.