Guanxi Huang
Research Subjects
Guanxi Huang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Pingze Zhang
College of Materials Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
2 shared publications - Qiang Yao
Jiangsu Product Quality Testing & Inspection Institute, Nanjing 210007, China
2 shared publications - Xiaolin Zhu
Jiangsu Product Quality Testing & Inspection Institute, Nanjing 210007, China
2 shared publications - Zhengjun Yao
College of Materials Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China
2 shared publications - Bo Liu
Jiangsu Product Quality Testing & Inspection Institute, Nanjing 210007, China
1 shared publication - Cairong Kang
Jiangsu Zhongxin Pipe Sci-Tec Co., Ltd., Nanjing 211100, China
1 shared publication
Latest Publications
- Open Access
Research on Automatic Identification and Rating of Ferrite–Pearlite Grain Boundaries Based on Deep Learning
Grain size has a significant effect on the mechanical properties of metals. It is very important to accurately rate the grain size number of steels. This paper presents a model for automatic detection and quantitative analysis of the grain size of ferrite–pearlite two-phase microstructure to segment ferrite grain boundaries. In view of the challenging problem of hidden grain boundaries in pearlite microstructure, the number of hidden grain boundaries is inferred by detecting them with the confidence of average grain size. The grain size number is then rated using the three-circle intercept procedure. The results show that grain boundaries can be accurately segmented by using this procedure. According to the rating results of grain size number of four types of ferrite–pearlite two-phase microstructure samples, the accuracy of this procedure is greater than 90%. The grain size rating results deviate from those calculated by experts using the manual intercept procedure by less than Grade 0.5—the allowable detection error specified in the standard. In addition, the detection time is shortened from 30 min of the manual intercept procedure to 2 s. The procedure presented in this paper allows automatic rating of grain size number of ferrite–pearlite microstructure, thereby effectively improving the detection efficiency and reducing the labor intensity.
Open AccessResearch on Automatic Identification and Rating of Ferrite–Pearlite Grain Boundaries Based on Deep Learning
Grain size has a significant effect on the mechanical properties of metals. It is very important to accurately rate the grain size number of steels. This paper presents a model for automatic detection and quantitative analysis of the grain size of ferrite–pearlite two-phase microstructure to segment ferrite grain boundaries. In view of the challenging problem of hidden grain boundaries in pearlite microstructure, the number of hidden grain boundaries is inferred by detecting them with the confidence of average grain size. The grain size number is then rated using the three-circle intercept procedure. The results show that grain boundaries can be accurately segmented by using this procedure. According to the rating results of grain size number of four types of ferrite–pearlite two-phase microstructure samples, the accuracy of this procedure is greater than 90%. The grain size rating results deviate from those calculated by experts using the manual intercept procedure by less than Grade 0.5—the allowable detection error specified in the standard. In addition, the detection time is shortened from 30 min of the manual intercept procedure to 2 s. The procedure presented in this paper allows automatic rating of grain size number of ferrite–pearlite microstructure, thereby effectively improving the detection efficiency and reducing the labor intensity.
- Open Access
- 19 February 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (2) , 482
Research on Intelligent Identification and Grading of Nonmetallic Inclusions in Steels Based on Deep Learning
Non-metallic inclusions are unavoidable defects in steel, and their type, quantity, size, and distribution have a great impact on the quality of steel. At present, non-metallic inclusions are mainly detected manually, which features high work intensity, low efficiency, proneness to misjudgment, and low consistency of results. In this paper, based on deep neural network algorithm, a small number of manually labeled, low-resolution metallographic images collected by optical microscopes are used as the dataset for intelligent boundary extraction, classification, and rating of non-metallic inclusions. The training datasets are cropped into those containing only a single non-metallic inclusion to reduce the interference of background information and improve the accuracy. To deal with the unbalanced distribution of each category of inclusions, the reweighting cross entropy loss and focal loss are respectively used as the category prediction loss and boundary prediction loss of the DeepLabv3+ semantic segmentation model. Finally, the length and width of the minimum enclosing rectangle of the segmented inclusions are measured to calculate the grade of inclusions. The resulting accuracy is 90.34% in segmentation and 90.35% in classification. As is verified, the model-based rating results are consistent with those of manual labeling. For a single sample, the detection time is reduced from 30 min to 15 s, significantly improving the detection efficiency.
- 19 February 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (2) , 482
Open AccessResearch on Intelligent Identification and Grading of Nonmetallic Inclusions in Steels Based on Deep Learning
Non-metallic inclusions are unavoidable defects in steel, and their type, quantity, size, and distribution have a great impact on the quality of steel. At present, non-metallic inclusions are mainly detected manually, which features high work intensity, low efficiency, proneness to misjudgment, and low consistency of results. In this paper, based on deep neural network algorithm, a small number of manually labeled, low-resolution metallographic images collected by optical microscopes are used as the dataset for intelligent boundary extraction, classification, and rating of non-metallic inclusions. The training datasets are cropped into those containing only a single non-metallic inclusion to reduce the interference of background information and improve the accuracy. To deal with the unbalanced distribution of each category of inclusions, the reweighting cross entropy loss and focal loss are respectively used as the category prediction loss and boundary prediction loss of the DeepLabv3+ semantic segmentation model. Finally, the length and width of the minimum enclosing rectangle of the segmented inclusions are measured to calculate the grade of inclusions. The resulting accuracy is 90.34% in segmentation and 90.35% in classification. As is verified, the model-based rating results are consistent with those of manual labeling. For a single sample, the detection time is reduced from 30 min to 15 s, significantly improving the detection efficiency.