Object-to-Text Conversion for Precise Positional Relation Description

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image recognition systems struggle to accurately describe the positional relations between objects in drawings, leading to variations in explaining these relations and difficulties in determining overlap states between objects, especially when using graphic drawing programs.

Innovation Solution

A method and system that convert objects into text by extracting coordinate data, determining positional relations, converting these into graph data, and generating text using an image processing unit, feature extraction unit, graph generation unit, and text generation unit, with a GUI for display.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If image recognition is used to describe positional relations between objects, then automation is improved, but measurement precision deteriorates due to variations in explaining positional relations

Engineering Contradiction:
Improveautomation of text generationVSAvoidprecision of positional relation description
Core Design Contradiction:
Extent of automationVSMeasurement precision

Solution Approach 1:

The patent introduces graph data as an intermediary representation between image recognition and natural language generation. The graph data structure with vertices and edges provides a standardized, precise way to represent positional relations, serving as a mediator that bridges the gap between automated image analysis and accurate textual description.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent transforms positional relation information from continuous spatial coordinates into discrete graph data parameters (vertices, edges, and their attributes). This parameter transformation enables precise representation of positional relations by converting vague spatial concepts into structured, computable data elements with defined properties.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If graphic drawing programs are used to display objects, then ease of operation is improved, but measurement precision deteriorates due to overlap ambiguities

Engineering Contradiction:
Improveease of drawing creationVSAvoidprecision of overlap state determination
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent uses graph data as an intermediary to resolve overlap ambiguities in graphic drawings. By representing objects and their spatial relationships in a structured graph format, the system can precisely determine overlap states without relying on visual interpretation of potentially ambiguous graphical displays.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the complex task of determining overlap states into discrete graph data processing steps. By breaking down the spatial relationship analysis into vertex and edge operations on graph structures, the system can systematically and precisely determine overlap states without the ambiguities inherent in visual graphic interpretation.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS12518449B2Object-to-text conversion method and system
Publication Date: 2026.01.06 SEMICON ENERGY LAB CO LTD
  • US12518449B2 patent drawing
  • US12518449B2 patent drawing
  • US12518449B2 patent drawing

AI summary

Text is generated from an object. Text is generated from a first object. The first object includes a second object and a third object. A step of detecting coordinate data of the second object is included. A step of detecting coordinate data of the third object is included. A step of extracting positional relation between the second object and the third object from coordinate data is included. A step of converting the extracted positional relation into graph data is included. A step of generating text about the positional relation between the second object and the third object from graph data is included.