A novel spatter detection algorithm based on typical cellular neural network operations for laser beam welding processes

Abstract

Real-time monitoring of laser beam welding (LBW) has increasingly gained importance in several manufacturing processes ranging from automobile production to precision mechanics. In the latter, a novel algorithm for the real-time detection of spatters was implemented in a camera based on cellular neural networks. The latter can be connected to the optics of commercially available laser machines leading to real-time monitoring of LBW processes at rates up to 15 kHz. Such high monitoring rates allow the integration of other image evaluation tasks such as the detection of the full penetration hole for real-time control of process parameters.

Multi-feature detection for quality assessment in laser beam welding: Experimental results

“…Simulation results were compared to those of a reference file which was built through visual inspection of the process images. Analyses of the detection rate of spatters and FPH were largely discussed in [14,15].…”

“…The proposed closed-loop control system leads to the on-line adjustment of the laser power with controlling rates of up to 14 kHz. Furthermore, in [14] we showed that the occurrence of spatters can also be observed in LBW in order to provide quality information about the process at monitoring rates of 15 kHz.…”

Monitoring of welding status by molten pool morphology during high-power disk laser welding

“…In this study, the molten pool around the keyhole was selected as the ROI (Region of Interesting, its resolution is 256 pixels and 211pixels) as it is the region which was the most violently varied part of molten pool, the cropped image is shown in Fig.5(a).The result of enhancement of Fig.5(a) with the piecewise linear stretching histogram equalization method is shown in Fig.5 The enhanced image shown in Fig.5 A c c e p t e d M a n u s c r i p t 11 areas, which were produced by spatters, also existed, and it will interfere the morphology information of molten pool. Through the experimental statistics, the area of these noisy regions was less than 200 pixels.…”

“…The molten pool formed due to integrated forces of the vapor pressure of liquid metal caused by high power laser beam, the surface tension of molten pool, the blowing force of shielding gas and the fluid gravity in the keyhole [11][12][13], and its morphology varied violently. Yamada et al carried out X-ray imaging of inside materials to observe behavior of molten pool using intense synchrotron radiation during laser welding [14], but the X-ray imaging system were expensive and harmful to operators.…”

NEROvideo: a general-purpose CNN-UM video processing system

“…Over the last 25 years cellular nonlinear networks (CNN) [10] have proved to be suitable for high-speed image processing in numerous fields of application [6,27,28,38]. The concept of the CNN universal machine (CNN-UM) [31] evolves these networks to freely programmable and universal processors which-in contrast to numerous specialised image processing machines-are not limited to certain operations, but offer a broad selection of image processing operators and algorithms [18].…”