Chemical molecular structure chart partition method based on area size and curvature

Abstract

The invention discloses a chemical molecular structure chart partition method based on area size and curvature. The partition method comprises the following steps: selecting a BMP (Bitmap) of a two-dimensional chemical molecular structure to be processed; performing image preprocessing, comprising operations such as binaryzation and denoising; adopting an eight-connected judgment algorithm to perform boundary following and marking each connected region; partitioning the chemical molecular structure map into two parts according to the size of the connected regions, wherein one part refers to the combination of all large-size connected regions composed of chemical bonds, and the other part refers to the combination of small-size regions such as heteroatoms, radical groups and chemical bonds; extracting single line segments and similar line segments representing the chemical bonds from the small-size combination diagram according to the curvature; at last, distinguishing the similar line segments and the single line segments according to the information such as location, combining the chemical bonds of the single line segments and the large-size connected regions to finish separation of the chemical bonds from the heteroatoms and the radical groups. The method can achieve good partition effect during image partitioning.

Description

technical field [0001] The invention belongs to the field of digital image processing, and in particular relates to a method for segmenting chemical molecular structure diagrams based on region size and tortuosity. Background technique [0002] chemical molecular structure Figure 1 Generally use ChemDraw, ISIS / Draw, DrawIt and ACD / ChemSketch software to draw. Drawing not only generates static chemical molecular structure pictures, but also translates the structural description information of the pictures into computer-readable file formats, such as ISIS, MOLfile, SMILES, which are called dynamic structural formula formats, and describe atoms, chemical bonds and connections between atoms. At present, molecular structure retrieval operations are generally based on dynamic structural formula formats, for example, "Design of Compound Structure Retrieval System and Its Implementation on the Uyghur Drug Resource Database", Feng Hongjun, Xinjiang University, 2008 master's thesis;...

AI Technical Summary

Problems solved by technology

From the perspective of pure images, there are few researches on the extraction of chemical molecular structure information and retrieval.

Images