AI Structural Formula Image Recognition via Adjacency Matrix
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Structural formula images in various literature and papers are difficult to search due to their image format, necessitating a method to convert them into a searchable string format.
Innovation Solution
An artificial intelligence device and method that recognizes structural formula images by obtaining information on atomic regions and bonding relationships, generating an adjacency matrix, and converting the image into a predetermined string format, such as SMILES.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If structural formula images are used to represent chemical structures, then geometric information and molecular structure details can be clearly displayed, but searching and retrieval become difficult
Solution Approach 1:
The patent creates a string representation (SMILES or InChI) as a copy of the structural formula image. This string format preserves the molecular structure information in a searchable text format, allowing users to search chemical structures without needing to process image data directly.
Solution Approach 2:
The patent transforms the visual parameters of the structural formula (positions, shapes, colors of atomic regions) into chemical parameters (atom types, bonding relationships). This parameter transformation converts image data into structured chemical information that can be searched and processed computationally.
2Ease of manufacture
If conventional image processing methods are used to recognize structural formulas, then implementation is simple, but bonding relationship recognition accuracy is insufficient
Solution Approach 1:
The patent divides the structural formula image into multiple atomic regions, each representing a specific atom. By segmenting the image and analyzing each region separately, the system can accurately identify atom types and their bonding relationships while maintaining implementation feasibility through modular processing.
3Measurement precision
If detailed analysis of each atomic region is performed to improve recognition accuracy, then bonding relationships can be identified more accurately, but processing time increases
Solution Approach 1:
The patent performs preliminary segmentation of the structural formula image into atomic regions before detailed analysis. This preliminary action organizes the data structure in advance, allowing subsequent bonding relationship analysis to be more efficient and accurate without excessive processing time.
Data Source
AI summary
An artificial intelligence device includes a memory in which a structural formula image is stored, and a processor configured to obtain information on a plurality of atomic regions from the structural formula image, to obtain information on bonding relationships between a plurality of atoms on the basis of the information on the plurality of atomic regions, to generate an adjacency matrix on the basis of the information on the plurality of atomic regions and the information on bonding relationships between the plurality of atoms, and to generate a predetermined string format corresponding to the structural formula image on the basis of the adjacency matrix.