• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.7
• /
• pp.865-870
• /
• 2018

A machine vision based industrial inspection includes defects detection and classification. Fast inspection is a fundamental problem for many applications of real-time vision systems. It requires little computation time and localizing defects robustly with high accuracy. Deep learning technique have been known not to be suitable for real-time applications. Recently a couple of fast region-based CNN algorithms for object detection are introduced, such as Faster R-CNN, and YOLOv2. We apply these methods for an industrial inspection problem. Three CNN based detection algorithms, VOV based CNN, Faster R-CNN, and YOLOv2, are experimented for defect detection on metal surface. The results for inspection time and various performance indices are compared and analysed.

• Design & Manufacturing
• /
• v.14 no.4
• /
• pp.52-57
• /
• 2020

In this study attempted to prevent defects due to blow holes among defects of sand casting products. It was intended to reduce the defect rate by reducing the blow hole of the inner surface. Currently, expectations and requirements for the quality level of non-ferrous aluminum casting in the casting industry are increasing. In addition, the shape is complex and the shrinkage precision is required. Among them, the test prototype is expensive to manufacture the mold, and the production time is also long, and the product is manufactured by sand casting. At this time, the highest defect rates are defects caused by shrinkage defects, surface defects, and blow holes.. At this study, the manufacturing time was shortened by using the shape of the fluid movement path in advance. Also, it is possible to reduce defects due to blow holes.

• Journal of Korea Society of Industrial Information Systems
• /
• v.17 no.2
• /
• pp.73-80
• /
• 2012

It has been highly required to develope an automatic metal surface inspection system, specifically using image processing techniques, which can replace the visual inspection method in the steel industry. In this paper, we propose a precisional surface measurement system using the moving average image processing technique. When the surface patterns which are generated in the rolling process of metal plates are recognized as defects, the proposed system can measure the actual number of defects. It has been proved that our system shows better results than the conventional FFT method.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.19 no.5
• /
• pp.1202-1210
• /
• 1995

Aiming at nondestructive evaluation of defect with high accuracy and resolution, ICFPD(Induced Current Focusing Potential Drop) technique was newly developed. This technique can be applied for locating and sizing of defects in components with not only simple shape such as plain surface but also more complex shape and geometry such as curved surface and dissimilar joing. This paper describes the principle of ICFPD technique and also the results of 2-dimensional surface crack in ferromagnetic metal(A508 Cl. III steel) and paramagnetic metal (pure aluminum and stainless 304 steel) measured by this technique. Results are that surface defects in each specimen are detected with the difference of potential drop, and potential drops are distributed a similar shape for each metal and each depth. The normalized potential drop ( $V_{\delta}$2/$^{t}$ / $V_{{\delta} 2}$$^{-1}$) max. in the vicinity of defect is varied with the depth of defect. Therefore, ICFPD technique can be used for the evaluation of defect not only in ferromagnetic metal but also in paramagnetic steel..

• Proceedings of the KWS Conference
• /
• 2002.10a
• /
• pp.402-407
• /
• 2002

Laminate tooling process is a fast and simple method to make metal tools directly for various molding processes such as injection molding in rapid prototyping field. Metal sheets are usually cut, stacked, aligned and joined with brazing or soldering. Through the joining process, all of the metal sheet layers should be rigidly joined. When joining process parameters are not appropriate, there would be defects in the layers. Among various types of defects, non-bonded gaps of the tool surface are of great importance, because they directly affect the surface quality and dimensional accuracy of the final products. If a laminate tool with defects has to be abandoned, it could lead to great loss of time and cost. Therefore a repair method for non-bonded gaps of the surface is essential and has important meaning for rapid prototyping. In this study, a rapid laminate tooling system composed of a CO2 laser, a furnace, and a milling machine was developed. Metal sheets were joined by furnace brazing, dip soldering and adhesive bonding. Joined laminate tools were machined by a high-speed milling machine to improve surface quality. Also, repair brazing and soldering methods of the laminates using the $CO_2$ laser system have been investigated. ill laser repair process, the beam duration, beam power and beam profile were of great importance, and their effects were simulated by [mite element methods. The simulation results were compared with the experimental ones, and optimal parameters for laser repair process were investigated.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.26 no.11
• /
• pp.1597-1604
• /
• 2002

In this study, numerical investigation has been performed on the impingement, spreading and solidification of a coating material droplet impacting onto a solid substrate in the thermal spray process. The numerical model is validated through the comparison of the present numerical result with experimental data fer the flat substrate without surface defects. An analysis of deposition formation on the non-polished substrate with surface defects is also performed. The parametric study is conducted with various surface defect sizes and shapes to examine the effect of surface defects on the impact and solidification of the liquid droplet on the substrate.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.02a
• /
• pp.61-61
• /
• 2010

In the present work, we present the optimized the hybrid structures of carbon nanotubes (CNTs) and metal nanocomposites including Cu, Al, Co and Ni using the first principle calculations based on the density functional theory. Introduction of CNTs into a metal matrix has been considered to improve the mechanical properties of the metal matrix. However, the binding energy between metals and pristine CNTs wall is known to be so small that the interfacial slip between CNTs and the matrix occurs at a relatively low external stress. The application of defective or functionalized CNTs has thus attracted great attention to enhance the interfacial strength of CNT/metal nanocomposites. Herein, we design the various hybrid structures of the single wall CNT/metal complexes and characterize the interaction between single wall CNTs and various metals such as Cu, Al, Co or Ni. First, differences in the binding energies or electronic structures of the CNT/metal complexes with the topological defects, such as the Stone-Wales and vacancy, are compared. Second, the characteristics of functionalized CNTs with various surface functional groups, such as -O, -COOH, -OH interacting with metals are investigated.We found that the binding energy can be enhanced by the surface functional group including oxygen since the oxygen atom can mediate and reinforce the interaction between carbon and metal. The binding energy is also greatly increased when it is absorbed on the defects of CNTs. These results strongly support the recent experimental work which suggested the oxygen on the interface playing an important role in the excellent mechanical properties of the CNT-Cu composite[1].

• Journal of Sensor Science and Technology
• /
• v.24 no.3
• /
• pp.169-174
• /
• 2015

In this study, we developed a system to detect the various defects in the metallic objects using the phenomenon that the defects cause the changes of the natural resonant frequencies. Our system consists of a FFT Amp, an Auto Impact Hammer, a Hammer controller and a PC. Auto Impact Hammer creates vibrations in the metallic objects when tapped on the surface. These vibrational signals are converted to the voltage signals by an acceleration sensor attached to the metallic part surface. These analog voltage signals were fed into an ADC (analog-digital converter) and an FFT (fast fourier transform) conversion in the FFT Amp to obtain the digital data in the frequency domain. Labview graphical program was used to process the digital data from th FFT amp to display the spectrum. We compared those spectra with the standard spectrum to find the shifts in the resonant frequencies of the metal parts, and thus detecting the defects. We used PCB's acceleration sensor and TI's TMS320F28335 DSP (digital signal processor) to obtain the resolution of 2.93 Hz and to analyze the frequencies up to 44 kHz.

• Journal of the Korean Society of Safety
• /
• v.10 no.4
• /
• pp.9-12
• /
• 1995

Detection of small defects by medical ultrasonic testing when the thermal sprayed coating by $MoSi_2$ on the metal is done. The defects may occur at the bonded surface. So, the detecting method of the defects by non-destructive in spection is desired. Here, in order to examine the possibility of the detection of the small defects by the ultrasonic. The electronic scanning ultrasonic equipment using an array probe developed as the medical ultrasonic diagnostic equipment is applied for the detection of the metal material defects. It's validity is investigated.

• The Journal of the Korea Contents Association
• /
• v.11 no.3
• /
• pp.18-25
• /
• 2011

In a manufacturing process of metal coins, surface defects of coins are manually detected. This paper describes an new method for detecting surface defects of metal coins on a moving conveyor belt using image processing. This method consists of multiple procedures: segmentation of a coin from the background, alignment of the coin to the model, projection of the aligned coin to the best eigen image space, and detection of defects by comparison of the projection error with an adaptive threshold. In these procedures, the alignement and the projection are newly developed in this paper for the detection of coin surface defects. For alignment, we use the histogram of the segmented coin, which converts two-dimensional image alignment to one-dimensional alignment. The projection reduces the intensity variation of the coin image caused by illumination and coin rotation change. For projection, we build multiple eigen image spaces and choose the best eigen space using estimated coin direction. Since each eigen space consists of a small number of eigen image vectors, we can implement the projection in real- time.