A method for predicting the corrosion degree of steel based on 3D optical profilometer to obtain point corrosion pit image

CN122199798APending Publication Date: 2026-06-12SHANGHAI UNIV

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Applications(China)
Current Assignee / Owner: SHANGHAI UNIV
Filing Date: 2026-03-10
Publication Date: 2026-06-12

AI Technical Summary

⚠Technical Problem

Existing technologies struggle to quickly and accurately obtain the three-dimensional morphological parameters of pitting corrosion in steel, leading to inaccurate assessments of corrosion severity and an inability to effectively reflect localized corrosion characteristics.

⚗Method used

A 3D optical profilometer was used to acquire the three-dimensional morphological data of pitting corrosion pits. The pits were then automatically detected and segmented using image recognition algorithms. Geometric and texture feature parameters were extracted to construct a corrosion degree prediction model.

🎯Benefits of technology

It enables quantitative and accurate assessment of steel corrosion, improves the accuracy and reliability of corrosion prediction, and features a high degree of automation, making it suitable for rapid testing in laboratories and engineering sites.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN122199798A_ABST

Patent Text Reader

Abstract

The application discloses a method for predicting the corrosion degree of steel based on a 3D optical profiler to obtain a pitting image, and belongs to the technical field of metal material corrosion monitoring and evaluation. The method first pretreats a steel sample, obtains a three-dimensional topographic image of the surface pitting of the steel sample through an interference scanning method of a 3D optical profiler, then performs segmentation processing on the image, and synchronously extracts geometric feature parameters and texture feature parameters of the pitting; based on the feature parameters and the known corrosion degree thereof, a sample data set is constructed, after data preprocessing, the construction and optimization of a machine learning model are completed through a core cycle including training, evaluation and hyperparameter optimization, and finally a corrosion degree prediction model is obtained. The feature parameters of a sample to be tested are input into the model, and the corrosion rate or corrosion grade thereof can be output. The application realizes rapid, accurate and quantitative evaluation of the corrosion degree of steel.

Need to check novelty before this filing date? Find Prior Art