نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Automatic inspection of steel surfaces is one of the basic processes in steel production. Most of steel producers have replaced the traditional human-based inspection methods with these new automatic and machine-based Methods. In this paper, a new approach has been proposed for detection and categorization of cold-rolled surface defects. The proposed algorithm is mainly based on artificial MLP neural networks and uncomplicated computational indicators. Experimental results show that the proposed method could detect up to 93.3% of prevalent defects. In addition, this method is appropriate for real-time implementation because of its satisfactory execution time on a modern computer.
کلیدواژهها English
