A Method for Quality Inspection of Continuous Casting Billet Based on Infrared Flaw Detection and 2-D Image

It is very important to improve product quality while reducing cost and increasing efficiency in steelmaking industry. The assurance of continuous casting billet quality is essential in the production process to ensure the manufacturing of high-quality products. This article introduces, for the first time, the combination of infrared thermal imaging and 2-D imaging into the steel manufacturing field, proposing a high-precision defect recognition method based on thermodynamics and lightweight networks. In addition, a scientific method for assessing the severity of defects in continuous casting billets has been proposed, aimed at evaluating their quality. The defect identification module obtains information on the location, size, and area of defects in the billets, while the evaluation module integrates this defect information and assigns a score. Comparative experiments with state-of-the-art (SOTA) models and production data verification demonstrate the superiority and effectiveness of the proposed method. This approach has achieved significant results in reducing costs and improving production efficiency.