In the continuous casting process, issues such as time lag in macroscopic inspection results and human interference factors hinder efficient quality assessment. Rapid and accurate identification of round bloom quality defects is critical for stabilising production. The study employs a U-Net neural network to preprocess image data from the continuous casting process, achieving a segmentation accuracy of 94.4% for round blooms. Metrics including precision, confusion matrix and ROC curve are used to evaluate the reliability and stability of different mathematical models in predicting defects. The convolutional neural network demonstrates superior performance with a defect identification accuracy of 97.12%, providing actionable guidance for on-site production optimisation.
LiuYWangXDSunY, et al.Research on a new detection method of slab surface crack in mould during continuous casting. Metall Etall Res Techol2018; 115: 82.
2.
ZhaoLMOuyangQChenDF, et al.Surface defects inspection method in hot slab continuous casting process. Ironmak Steelmak2011; 38: 464–470.
3.
KonovalenkoIVMarushchakPOKuzON. A method for processing and analysis of the images of a network of thermal fatigue cracks on the surfaces of rollers of continuous casting machines. Mater Sci2018; 54: 175–183.
4.
MengXLLuoSWangWL, et al.A detection model for corner cracks of continuous casting strand based on deep learning. Ironmak Steelmak2022; 49: 1048–1056.
5.
WangYYWangQCChengYH, et al.Investigation on mold breakout prediction methods based on image recognition via convolutional neural network. Metall Mater Trans B2024; 55: 2685–2698.
6.
ZhangLQiaoZLLiL. An evolutionary deep learning method based on improved heap-based optimization for medical image classification and diagnosis. IEEE Acces2024; 12: 102745–102773.
7.
ZhuNYLiuXLiuZQ, et al.Deep learning for smart agriculture: concepts, tools, applications, and opportunities. Int J Agric Biol Eng2018; 11: 13.
8.
XuCJDuYZhengWJ, et al.Facial expression recognition based on YOLOv8 deep learning in complex scenes. Int J Inf Commun Technol2025; 26: 89–101.
9.
ChuryumovAYKazakovaAA. Prediction of true stress at hot deformation of high manganese steel by artificial neural network modeling. Materials2023; 16: 1083.
10.
HonyszRSoria, BeatrizL, et al.Modeling the chemical composition of ferritic stainless steels with the use of artificial neural networks. Metals2021; 11: 724.
11.
ChuryumovAKazakovaAChuryumovaT. Modelling of the steel high-temperature deformation behaviour using artificial neural network. Metals2022; 12: 447.
12.
GengMYMaHAWangJX, et al.A deep learning framework for predicting slab transverse crack using multivariate LSTM-FCN in continuous casting. ESWA2025; 260: 125413.
13.
LiYWangWLZhouLJ, et al.Prediction and feature analysis of entrapped slag defect on casting slab based on optimized XGBoost model. Metall Mater Trans B2024; 55: 2026–2036.
14.
YeXMWuXGuoY. Real-time quality prediction of casting billet based on random forest algorithm. IEEE Acces2018: 140–143.
15.
YangHLFangYMLiuL, et al.Improved YOLOv5 based on feature fusion and attention mechanism and its application in continuous casting slab detection. IEEE TIM2023; 72: 1–16.
16.
LianGHSunQHLiuXM, et al.Automatic recognition and intelligent analysis of central shrinkage defects of continuous casting billets based on deep learning. J Iron Steel Res Int2023; 30: 937–948.
17.
LinJHYaoYMaL, et al.Detection of a casting defect tracked by deep convolution neural network. Int J Adv Manuf Tech2018; 97: 573–581.
18.
JuHFangYMYangHL, et al.Improved lightweight YOLOv8 with DSConv and reparameterization for continuous casting slab detection on embedded device. IEEE TIM2025; 74: 3509532.
19.
BeecheCSinghJPLeaderJK, et al.Super U-Net: a modularized generalizable architecture. PR2022; 128: 108669.
20.
GongDQYangJYGanXY, et al.Fully automated construction of a deep U-net network model for medical image segmentation. Sensor Mater2023; 35: 4633–4652.
21.
SenapatiPBasuADebM, et al.Sharp dense U-Net: an enhanced dense U-Net architecture for nucleus segmentation. IJMLC2024; 15: 2079–2094.
22.
YinHTShaoYD. CFU-Net: a coarse–fine U-Net with multilevel attention for medical image segmentation. IEEE TIM2023; 72: 1–12.
23.
HanNZhouLXieZM, et al.Multi-level U-net network for image super-resolution reconstruction. Displays2022; 73: 102192.
24.
RajamaniKTRaniPSiebertH, et al.Attention-augmented U-Net (AA-U-Net) for semantic segmentation. SIVP2023; 17: 981–989.
25.
PengRBVarshneyPK. On performance limits of image segmentation algorithms. CVIU2015; 132: 24–38.
26.
YuLTLiZBXuM, et al.Distribution-Aware margin calibration for semantic segmentation in images. IJCV2022; 130: 95–110.
27.
LuKWangWYPanXJ, et al.Resformer-Unet: a U-shaped framework combining ResNet and transformer for segmentation of strip steel surface defects. ISIJ Int2024; 64: 67–75.
28.
ZhangDJHeFZHanS, et al.An efficient approach to directly compute the exact Hausdorff distance for 3D point sets. Integr Comput-aid E2017; 24: 261–277.
29.
ElanangaiVVasanthK. An automated steel plates fault diagnosis system using adaptive faster region convolutional neural network. JIFS2022; 43: 7067–7079.
30.
ZengWHWangJYChenP, et al.Surface defect identification method for hot-rolled steel plates based on random data balancing and lightweight convolutional neural network. Signal Image Video P2024; 18: 5775–5786.
31.
HuangCYDaiZJSunY, et al.Recognition of converter steelmaking state based on convolutional recurrent neural networks. Metall Mater Trans B2024; 55: 1856–1868.
32.
HeDXuKZhouP. Defect detection of hot rolled steels with a new object detection framework called classification priority network. CIE2019; 128: 290–297.
33.
LisK. An Analysis of the effect of skew rolling parameters on the surface quality of C60 steel parts using classification models. Materials2024; 17: 5362.
34.
WangYSDingYShahrampourS. TAKDE: temporal adaptive kernel density estimator for real-time dynamic density estimation. IEE IPAMI2023; 45: 13831–13843.
