Semantic Image Navigation for Large Entities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing photo tours using user photographs for landmarks lack informative content beyond the landmark's name and cannot effectively represent entities that cannot be captured by a single photograph, such as large structures.
Innovation Solution
A method for generating semantic image navigation experiences by selecting entities, identifying sub-entities with containment relationships, filtering and ranking them based on landmark popularity, and creating a sequence of images from pre-stored navigation experiences for display, including information about each sub-entity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If photo tours show one entity from a variety of viewpoints, then the visualization is coherent and informative, but the information content remains limited to the landmark name
Solution Approach 1:
The patent segments a large entity (e.g., a city or region) into multiple sub-entities (landmarks, points of interest), each with its own photo tour. This allows the system to provide rich information content by combining multiple focused tours rather than attempting to cover everything in a single tour, thereby reducing the information loss without requiring excessive system complexity.
Solution Approach 2:
The patent implements a nested structure where sub-entity photo tours are nested within a parent entity photo tour. The parent tour provides an overview and can include or link to detailed sub-entity tours, creating a hierarchical information structure that delivers comprehensive content while maintaining organizational simplicity.
2Adaptability or versatility
If photo tours are created for entities that cannot be captured easily by a user photograph, then coverage is improved, but the ability to create tours is limited
Solution Approach 1:
The patent uses automated image recognition and processing to create photo tours without requiring manual photographing by users. The system processes existing images and automatically generates navigation experiences, making tour creation accessible for entities that cannot be easily captured by simple user photographs, such as large structures or complex scenes.
Solution Approach 2:
The system performs automatic entity recognition, image selection, and tour generation without requiring manual user input for each tour creation. This self-service approach enables the system to handle diverse entities autonomously, significantly improving adaptability while maintaining ease of manufacture.
3Loss of information
If multiple sub-entities are included in a navigation experience, then comprehensiveness is improved, but the navigation experience becomes more complex
Solution Approach 1:
The patent divides a complex navigation experience into multiple manageable sub-entity tours, each focusing on a specific landmark or point of interest. This segmentation allows the system to present comprehensive information about multiple sub-entities while keeping each individual tour segment simple and easy to navigate, thereby reducing overall complexity.
Solution Approach 2:
The patent organizes multiple sub-entity tours in a hierarchical structure with a parent entity tour at one level and sub-entity tours at another level. This dimensional organization allows users to navigate from general to specific information, making the comprehensive experience more manageable and less complex by adding a structural dimension to the navigation.
Data Source
AI summary
Aspects of the disclosure relate to generating a sequence of images or other visual representations associated with an entity, otherwise known as a semantic image navigation experience. After an entity is selected, a set of sub-entities may be identified. Each sub-entity in the set has a containment relationship with the selected entity as well as at least one associated landmark and one associated pre-stored navigation experience. Then, a ranking order of the sub-entities in the set may be determined based on characteristics of each entity. Based on the determined ranking order, a subset of sub-entities may be selected. A semantic image navigation experience for the selected entity may then be generated using the pre-stored navigation experiences associated with the subset of sub-entities.