Image Search System Using Object Positional Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for reviewing conference scenes in images fail to efficiently search for specific scenes based on the positional relationships of objects and persons in real spaces, requiring users to remember the layout of the conference room and taking significant time to find desired information as context information increases.
Innovation Solution
An information processing system and method that stores image information of a given space with positional information of objects, allowing users to search for scenes by specifying objects' positions, using a memory portion, receiving portion, and search portion to narrow down the scene based on the positional relationship of objects like persons or things.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional image searching methods are used that store context information of conferences, then users can access detailed context information, but it takes much time to search for the desired information as context information increases
Solution Approach 1:
The patent segments the search process by introducing intermediate search results (thumbnails with time stamps) between the initial search query and the detailed context information. This segmentation allows users to quickly filter through search results using visual cues before accessing full details, significantly reducing search time while maintaining information accessibility.
Solution Approach 2:
The patent adds a visual dimension to the search results by displaying thumbnails alongside textual context information. This dimensional enhancement allows users to quickly assess relevance through visual cues rather than reading through all textual information, thereby reducing search time without losing access to detailed context.
2Ease of operation
If users select conference rooms and arrange persons and objects based on standard layouts, then the search method can be utilized, but users cannot search effectively if they cannot remember the layout of the conference room
Solution Approach 1:
The patent creates visual copies (thumbnails) of the actual conference scenes as they appeared during the meeting. These thumbnails serve as visual replicas that users can recognize without needing to remember the original layout, allowing effective search based on visual memory rather than spatial memory of the conference room arrangement.
3Reliability
If detailed context information is stored for comprehensive search coverage, then all positional relationships can be searched, but the amount of information to be processed increases significantly
Solution Approach 1:
The patent segments information into two tiers: essential visual information (thumbnails) and detailed context information. This segmentation allows the system to maintain comprehensive search coverage through detailed information while reducing processing complexity by using simplified visual representations for initial filtering and matching.
Solution Approach 2:
The patent implements partial action by storing and processing only the most visually distinctive and relevant portions of conference scenes as thumbnails, rather than processing all possible positional relationship data. This partial processing approach maintains reliable search coverage for key elements while significantly reducing overall information processing complexity.
Data Source
AI summary
An information processing system includes a memory portion that stores each scene of image information of a given space captured by an image capturing portion in association with positional information on an object existent in the given space, a receiving portion that receives information that specifies the object as a search request, and a search portion that searches the image information stored in the memory portion for a scene where the object is existent in a given position on the basis of the search request.