The detection and classification method of sub-macroscopic defects inside steel with an ultrasonic testing and CatBoost-based stacking model

Abstract

With the increasing demand for advanced steel is increasing year by year, and the internal cleanness content of steel inclusions becomesis an important evaluation indicator for the evaluation of material material quality. Sub-macroscopicInclusions defects are randomly distributed inside the steel materials, which has a great impact on the performance and quality safety of the steel. In especial, sub-macroscopic inclusions with sizes ranging from 50μm to 400μm have seriously affected material stability and fatigue life because they are not covered by existing testing standards. In addition,Besides, the existing current detection methods for inclusions in steel generally have problems such as low efficiency and complexity process. In this paper, we propose a non-destructive inclusion testing and classification framework basing on ultrasonic testing experiments, signal feature extraction and machine learning methods. Under the optimal experimental detection conditions we found through experiments, a large-scale sub-macroscopic inclusion signal data set is established to realize the classification of defects. Moreover, Empirical Mode Decomposition (EMD) and other feature extraction algorithms are applied to further boost the model performance. We propose a Catboost-based stacking fused model named Stacked-CBT, which obtains state-of-the-art experimental result with accuracy rate of 86.65% and demonstrates that the proposed framework is feasible to classify the sub-macroscopic inclusion signals. To the best of our knowledge, there is no previous study in this field has acquire such large amount of experimental sub-macroscopic signal data while taking into consideration classification-specific designs.

Fabric defect detection based on feature enhancement and complementary neighboring information

“…Traditional fabric defect detection methods rely on image processing techniques to extract texture features from fabrics and perform quantitative and qualitative descriptions to achieve defect detection. These traditional detection algorithms mainly fall into two types: 1) Frequency domain algorithms: Common frequency domain algorithms include the Fourier Transform method [3], Gabor filtering [4], and Wavelet Transform [5], among others. These algorithms are suitable for detecting defects in solid-color fabrics with simple background features.…”