Melt pool segmentation for additive manufacturing: A generative adversarial network approach. (June 2021)

• Record Type: Journal Article
• Authors: Liu, Weibo; Wang, Zidong; Tian, Lulu; Lauria, Stanislao; Liu, Xiaohui
• Is Part Of: Computers & electrical engineering. Volume 92(2021)
• Publication Date: 2021-06
• Subjects: Additive manufacturing -- Generative adversarial network -- Defect detection -- Image processing -- Image segmentation -- Thermal image; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107183
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: Additive manufacturing (AM) is a popular manufacturing technique which is broadly exploited in rapid prototyping and fabricating components with complex geometries. To ensure the stability of the AM process, it is of critical importance to obtain high-quality thermal images by using image processing techniques. In this paper, a novel image processing method is put forward with aim to improve the contrast ratio of the thermal images for image segmentation. To be specific, an image-enhancement generative adversarial network (IEGAN) is developed, where a new objective function is designed for the training process. To verify the superiority and feasibility of the proposed IEGAN, the thermal images captured from an AM process are utilized for image segmentation. Experiment results demonstrate that the developed IEGAN outperforms the original GAN in improving the contrast ratio of the thermal images. Graphical abstract: Highlights: An image enhancement generative adversarial network is proposed for improving the contrast ratio of the image. A designed penalty term is introduced in the loss function of the generative adversarial network. The contrast improvement index is utilized for performance evaluation. The proposed algorithm is successfully employed in thermal imaging analysis.