Automation of geometric feature computation through image processing approach for single-layer laser deposition process. Issue 9 (1st September 2020)

• Record Type: Journal Article
• Authors: Patil, Deepika Bhanudas; Nigam, Akriti; Mohapatra, Subrajeet
• Is Part Of: International journal of computer integrated manufacturing. Volume 33:Issue 9(2020)
• Page Start: 895
• Page End: 910
• Publication Date: 2020-09-01
• Subjects: Laser deposition -- image processing -- image recognition -- feature extraction -- curve fitting -- inflexion points; Computer integrated manufacturing systems -- Periodicals; 670.427
• DOI: 10.1080/0951192X.2020.1815843
• Languages: English
• ISSNs: 0951-192X

Abstract

ABSTRACT: This paper reports the development of an image processing approach that can automatically compute the geometric features of the single layer deposition of Ti-6Al-4 V on the same substrate material using laser deposition process. It involves (i) conducting experiments and capturing the images of deposition geometry using scanning electron microscopy (SEM), (ii) development of the image processing technique to accurately parameterize the regular/irregular shape and size (i.e. width, height and area of deposition) of the single-layer deposition geometry and (iii) compare the accuracy of the proposed image processing technique with previously used methods (i.e. manual measurement and curve fitting) to obtain dimensions of the deposition geometry. Comparing the accuracy of the proposed image processing approach, against the manually measured width and height of deposition revealed an average relative error of 1% and 1.255% respectively. Comparing the area of deposition geometry obtained by curve fitting with manually measurement and the proposed image processing approach revealed that the maximum relative error between them is 9.5% and 5.84% respectively. Therefore, the proposed image processing approach demonstrates as a viable technique to compute the shape and size of the actual regular/irregular deposition geometry with sufficiently high accuracy and less human interference.