Non-metallic inclusion full-view-field quantitative statistical distribution characterization method

Abstract

The invention discloses a non-metallic inclusion full-view-field quantitative statistical distribution characterization method. The method comprises the steps that S1, establishing an image database composed of non-metallic inclusions, scratches and external interference images; s2, establishing a target detection model based on the non-metallic inclusion characteristic spectrum; s3, carrying outfull-view-field non-metallic inclusion characteristic data automatic collection on the processed surface of the to-be-measured metal material; s4, segmenting and extracting the non-metallic inclusionsin a manner of being accurate to a pixel level; and S5, in-situ quantitative statistical distribution characterization of large-range full-view-field non-metallic inclusions. According to the non-metallic inclusion full-view-field quantitative statistical distribution characterization method provided by the invention, by establishing the target detection model, the non-metallic inclusions in themetal material are subjected to full-view-field recognition and positioning, so that the recognition accuracy can be improved, errors caused by manual recognition are avoided, and the automation levelis relatively high.

Description

technical field [0001] The invention relates to the technical field of detection of non-metallic inclusions in metal materials, in particular to a quantitative statistical distribution characterization method of non-metallic inclusions in a full field of view. Background technique [0002] Non-metallic inclusions are divided into endogenous and exogenous inclusions. The oxides formed by metal deoxidation, the sulfides and nitrides precipitated when the solubility of sulfur and nitrogen are reduced during the solidification of steel are collectively referred to as endogenous inclusions, and the refractories introduced by metals in the smelting process , slag, etc. are foreign inclusions. It is generally believed that the quantity, size, composition and distribution of non-metallic inclusions are one of the important factors affecting the properties of steel. The existence of non-metallic inclusions blocks the continuity of the metal matrix, so that the stress between the mat...

AI Technical Summary

Problems solved by technology

This method needs to manually identify the types of inclusions in the field of view, and the statistical field of view is single. It takes time and effort to inspect the entire polished surface, and there is a problem of low efficiency. It cannot quantify the size distribution of inclusions on the entire polished surface, nor can it reflect the entire polished surface. Due to the difference in inclusion distribution, it is impossible to quickly realize the statistical quantitative distribution of non-metallic inclusions on the material surface

In addition, metallography, Image-Pro Plus and other software are often used to automatically identify the type of inclusions and quantitatively count the inclusion area. The software has high requirements for image quality and still needs to be supplemented by manual operation, so there are certain limitations.

In summary, the traditional non-metallic inclusion determination method can no longer meet the needs of material workers to investigate the process and improve material properties

Images