AI Polymer Structure Recognition for Copolymer Formula Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately recognize and identify polymer molecular structure formulas, particularly those of copolymers, due to inconsistencies in two-dimensional representations, making it difficult to utilize machine learning for polymer property prediction and synthesis.
Innovation Solution
A system utilizing artificial intelligence (AI) with a first model to detect bracket pairs and subscripts, and a second model to provide group information, enabling accurate recognition of copolymer molecular structures by converting images into machine-readable formats like SMILES.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single model is used to detect polymer molecular structure formulas, then the device complexity is reduced, but the measurement precision and reliability of recognizing copolymer structures deteriorate
Solution Approach 1:
The system divides the recognition task into two separate models: a first model that detects bracket pairs and subscripts, and a second model that provides group information about these elements. This segmentation allows each model to specialize in specific aspects of polymer structure recognition, improving overall precision while maintaining manageable complexity through modular architecture
Solution Approach 2:
The dual-model system creates a universal recognition framework that handles both simple and complex polymer structures (including copolymers) through complementary functions. The first model provides structural detection while the second model provides grouping context, together enabling versatile recognition across different polymer types without requiring separate specialized systems
2Ease of operation
If traditional search methods are used to identify polymer molecular structures, then the ease of operation is maintained, but the productivity and measurement precision deteriorate due to inability to consistently identify structures from two-dimensional images
Solution Approach 1:
The system replaces traditional manual search and visual inspection methods with an automated AI-based recognition system. The artificial intelligence models automatically detect and interpret polymer molecular structures from two-dimensional images, eliminating the need for manual searching while significantly improving both efficiency and consistency of structure identification
Solution Approach 2:
The recognition system performs self-service by automatically detecting bracket pairs, subscripts, and group information directly from input images without requiring manual annotation or intervention. The models process the images autonomously, extracting structural information and generating polymer identifiers independently, thereby大幅提高 productivity while maintaining ease of use
3Measurement precision
If manual methods are used to process polymer molecular structure formulas, then the measurement precision might be maintained, but the productivity and loss of time worsen due to difficulty in consistently identifying structures throughout chemical literature
Solution Approach 1:
The system performs preliminary detection of bracket pairs and subscripts before final structure identification. The first model pre-processes the image by locating and detecting these critical structural elements, preparing the data for the second model to interpret. This preliminary action enables faster processing while maintaining accuracy by focusing computational resources on key structural features
Data Source
AI summary
A system, method, and program for recognizing a polymer molecular structure formula using artificial intelligence are disclosed. The system includes at least one processor, and at least one memory storing a command or information that causes the at least one processor to perform an operation, wherein the operation includes detecting a polymer molecular structure formula image to generate detection data, the detection data including information about atomic regions including atoms, bonding between the atoms, and a bracket pair with an associated subscript, the system further including inputting the detection data to each of a first model and a second model to output first cluster data from the first model and to output from the second model second cluster data including group information about the bracket pair and the associated subscript and including information different from the first cluster data.