Transferring Metadata from Panoramic Imagery to Matched Photos

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current online map systems face challenges in keeping panoramic imagery up-to-date due to environmental changes, as uploaded user images lack metadata from the original panoramic imagery, leading to outdated information.

Innovation Solution

A system that uses computer vision techniques to match uploaded images with panoramic imagery, transferring metadata from the panoramic images to the user-uploaded images by aligning features and associating geolocation data, allowing for the annotation of user images with relevant metadata.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If panoramic imagery is updated periodically through expensive stitching processes, then metadata accuracy is maintained, but update frequency is limited and cost increases

Engineering Contradiction:
Improvemetadata accuracyVSAvoidupdate frequency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent transfers metadata from panoramic imagery to user-uploaded images by matching features between the two. Instead of re-stitching panoramas to update metadata, the system copies metadata attributes from the panoramic image to the uploaded image that has been matched to it, enabling frequent updates without expensive processing

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The system allows users to upload their own images which are then automatically matched to panoramic imagery and annotated with metadata. The uploaded images themselves serve as the update mechanism, eliminating the need for systematic re-acquisition and re-stitching of panoramic data

Inventive Principle:
Principle #25Self-service

2Reliability

If users upload recent images to replace outdated panoramic imagery, then imagery currency is improved, but metadata is lost

Engineering Contradiction:
Improveimagery currencyVSAvoidmetadata
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The core function of the system is to copy metadata from the panoramic image to the user-uploaded image. The metadata module extracts metadata attributes from the matched panoramic image and associates them with the uploaded image, preserving all marker information while using the current uploaded imagery

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The system merges the currency of user-uploaded images with the metadata richness of panoramic imagery. By annotating uploaded images with transferred metadata, the system combines the advantages of both approaches: current imagery from users and comprehensive metadata from the panoramic system

Inventive Principle:
Principle #5Merging (Combining)

3Manufacturing precision

If panoramic imagery is systematically gathered through large-scale efforts, then image quality is high, but processing time and cost increase

Engineering Contradiction:
Improveimage qualityVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

Instead of systematically gathering and processing all panoramic imagery, the system processes only the specific portion of the panorama that matches the uploaded image. The assignment module identifies and processes only the relevant matched region, extracting and transferring metadata only for that specific area rather than the entire panorama

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS8933929B1Transfer of annotations from panaromic imagery to matched photos
Publication Date: 2015.01.13 GOOGLE LLC
  • US8933929B1 patent drawing
  • US8933929B1 patent drawing
  • US8933929B1 patent drawing

AI summary

Systems and methods are disclosed for transferring information metadata from a first digital image to a second digital image. In one embodiment, an assignment module is configured to assign a corresponding portion of the first image to the second image using geolocation data. An extraction module is configured to extract a collection of features associated with the second image and the corresponding portion of the first image. An alignment module is configured to align the second image with a portion of the first image by transforming the second image so that features associated with the second image are geometrically aligned with the corresponding features of the portion of the first image. A metadata module is configured to associate metadata from the portion of the first image with the transformed second image. An annotation module is configured to annotate the second image with the associated metadata to generate an annotated image.