Naslov
Enhancing IR Thermographic Inspection of Subsurface Defects by Using the Technique of Edge Detection

Autori
Bagavac, Petra ; Krstulović-Opara, Lovre ; Domazet, Željko

Izvornik
Russian journal of nondestructive testing (1061-8309) 57 (2021), 7; 609-618

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
data filtering ; edge detection ; infrared thermography ; pre-impregnated CFRP composite plates

Sažetak
This paper presents a novel method for separating sub-surface defects in cured pre- impregnated carbon fiber reinforced polimer specimens on thermal images based on edge detection. Every specimen was recorded by means of infrared thermography in reflective configuration and raw data was processed by several algorithms: Thermal Signal Reconstruction, Fast Fourier Transformation and Principal Component Analysis. Images were processed to determine damaged and non-damaged areas. Algorithm for determining area of discontinuity is based on edge detection techniques. Edge detection enables both simulated damages and boundaries between specimen and the background in the thermal image to be detected. Proposed algorithm is emphasizing edges within Sobel filter. Edges are defined as transition zone between damaged and non-damaged material. In this article edge detection algorithm is used as decision assistance for non-destructive evaluation.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo

POVEZANOST RADA