Review of Research on Ceramic Surface Defect Detection Based on Deep Learning

The detection of ceramic surface defects is of great significance for product quality. Traditional detection methods have limitations, while deep learning methods bring new opportunities. This article first introduces the basic steps and current situation of data preparation. Secondly, it explores the imbalanced sample problem faced in ceramic surface defect detection based on methods such as data augmentation, sample distribution optimization, network structure improvement, and loss function design. It also reviews the small sample problem in ceramic surface defect detection through methods like data augmentation, transfer learning, unsupervised learning, and network structure optimization. The methods to improve the detection accuracy of small target defects on ceramic surfaces are elaborated, including adding attention mechanisms, feature improvement, network structure optimization, etc. The improvement of the real - time performance of model defect detection is analyzed from two aspects: the improvement of lightweight models and the integration and optimization of network modules. Finally, the solutions that can be used in the implementation of ceramic surface defect detection technology are summarized, and the future research directions of ceramic surface defect detection are prospected.