|
| Author/Year | Problem | Solution |
|
| Ye and Toyama [6] | To analyze the efficiency of various deep learning architectures. | A total of 7000 real-time images are evaluated. A total of 17 flaws are detected. |
| Ajmi et al. [8] | Defect detection and classification on small weld X-ray image datasets. | Data augmentation and deep learning techniques utilized for obtaining best results. |
| Mery [7] | To automate the process of defect detection in aluminum castings. | Convolution neural network (CNN) model utilized for effective detection of defects |
| Daniel et al. [1] | Internal defects in pipes (0 to 2 inches). | Vertical insertion camera. |
| Xiao et al. [9] | To detect weld bead width and depth of penetration defects in welds. | Coaxial infrared pyrometer |
| Schaunberger et al. [10] | To identify weld seam defects such as pores, tapers, and regressions (copper) | Defect detected from temperature curves. |
| Gao et al. [11] | Process stability and weld formation (laser welding) | Analysis with a high-speed camera. |
| Lei et al. [12] | Influence of thermal effect on droplet transfer (cold metal transfer (CMT) laser welding) | Analysis with a high-speed camera and brightness curves |
| Huang et al. [13] | Welding defect identification (laser welding) | The defect identification through electrical signals of laser plasma and plasma flumes acquired by a high-speed camera |
| Gao et al. [14] | To identify invisible weld defects | Magneto-optical imaging system and grayscale curves |
| Hamade and Baydoun [4] | To identify wormhole defects in welded lap joints. | X-ray computed tomography (CT) scan and Otsu segmentation |
| Zhang et al. [2] | To detect weld seam penetration defects | Multiangle image acquisition and convolution neural network (CNN) |
| Jiang et al. [5] | To identify porosity defects at ambient pressures. | High-speed camera. No defects under vacuum |
| Xie et al. [15] | To detect metal rust | High-speed images and Acoustic emission signal of pulsed laser |
| Zhou et al. [16] | To identify surface pit, spatter, softening in heat-affected zone (HAZ), oxide, and porosity | Addition of Sn foil |
| Choi et al. [17] | To detect lack of fusion (LOF), gas pores | Laser metal deposition technique and fatigue test to check efficiency. |
| Bacioiu et al. [18] | To monitor tungsten inert gas welding | High dynamic range camera and Fully convolution neural networks and convolution neural networks (FCN & CNN). |
| Shah and Liu [19] | To identify interfacial cracks, solidification cracks, surface defects, and oxides | Ultrasonic waves in resistance spot welding (URSW) |
| Nacereddine et al. [20] | To detect cracks, porosity, lack of penetration (LOP), and solid inclusion | Classification in radiographic images. |
| Francis et al. [21] | To analyze the potential of vacuum laser welding for thicker areas of nuclear parts. | Achieves the required weld quality equivalent to the electronbeam welding (EBW). |
| Zhang et al. [22] | Comprehensive insights of laser welding process. | Multiple optical sensor systems |
| Xu et al. [23] | To identify Keyhole-induced porosity | Three-dimensional transient model |
| Chaoudhuri et al. [3] | To identify Inherent flaws and fatigue cracks | Stress analysis and micro-computed tomography (CT) |
| Son et al. [24] | To examine the strength that exists between a material deposited and its substrate. | High bonding strength verified through Shear tests. |
| Reisgen et al. [25] | To detect porosity defects | )Nonvacuum electron beam welding (NV-EBW) |
| Millon et al. [26] | To identify the lack of fusion (LOF) or porosity defects | Laser ultrasonic signals. |
| Wu et al. [27] | Expulsion identification | Welding force signal |
| Xie et al. [28] | Heat-affected zone (HAZ) Cracks(liquidation and strain age cracks) | Postweld hot isostatic pressing |
| Qian et al. [29] | To identify high residual stress | Spontaneous magnetic signals |
| Kim et al. [30] | To detect welding defects in underground curled pipelines | Magnetic flux leakage (MFL) sensor signals |
| Hongmin and Wang [31] | To identify tiny weld bead flaws (cracks, pores, lack of fusion (LOF), cavities) | Closed magnetic reluctance signals |
| Zhang et al. [32] | To detect flaws in power disk laser welding | Spectrometer signals |
| Li and Lu [33] | To fetch a novel alloy for Biomedical utilization with the apt Young’s modulus | Powder metallurgy procedure is employed. |
| Proposed system | To identify porosity, crack, internal defects, and corrosion | Thermal images and fuzzy deep learning algorithm |
|
