• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.35.3-35
• /
• 2001

Developed is an optical auto-inspection system to detect some microscopic defects on the Inside surface of the hydraulic automobile brakes at the production line. A small cylindrical detection module with a solid laser source at its center has two rings of optical fibers to separately collect light reflected and scattered from the defects on the surface. The cylindrical brake part rotates with respect to the detection module that will move parallel to the rotational axis of the cylinder. Thus, the optical module can scan the whole inside surface area. The automatic detection of the defects is to compare the intensity distributions ...

• 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.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.7
• /
• pp.1100-1104
• /
• 2017

A detector-based approach shows the limited performances for the defect inspections such as shallow fine cracks and indistinguishable defects from background. Deep learning technique is widely used for object recognition and it's applications to detect defects have been gradually attempted. Deep learning requires huge scale of learning data, but acquisition of data can be limited in some industrial application. The possibility of applying CNN which is one of the deep learning approaches for surface defect inspection is investigated for industrial parts whose detection difficulty is challenging and learning data is not sufficient. VOV is adopted for pre-processing and to obtain a resonable number of ROIs for a data augmentation. Then CNN method is applied for the classification. Three CNN networks, AlexNet, VGGNet, and mofified VGGNet are compared for experiments of defects detection.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1996.11a
• /
• pp.444-449
• /
• 1996

The semiconductor, which is precision product, requires many inspection processes. The surface conditions of the semiconductor chip affect on the functions of the semiconductors. The defects of the chip surface are cracks or voids. As general inspection method requires many inspection procedure, the inspection system which searches immediately and precisely the defects of the semiconductor chip surface is required. We suggest the detection algorithm for inspecting the surface defects of the semiconductor surface. The proposed algorithm first regards the semiconductor surface as random texture and point spread function, and secondly presents the character of texture by linear estimation theorem. This paper assumes that the gray level of each pixel of an image is estimated from a weighted sum of gray levels of its neighbor pixels by linear estimation theorem. The weight coefficients are determined so that the mean square error is minimized. The obtained estimation window(two-dimensional estimation window) characterizes the surface texture of semiconductor and is used to discriminate the defects of semiconductor surface.

• International Journal of Railway
• /
• v.1 no.1
• /
• pp.1-5
• /
• 2008

The railway system requires safety and reliability of service of all railway vehicles. Suitable technical systems and working methods adapted to it, which meet the requirements on safety and good order of traffic, should be maintained. For detection of defects, non-destructive testing methods-which should be quick, reliable and cost-effective - are most often used. Since failure in railway wheelset can cause a disaster, regular inspection of defects in wheels and axles are mandatory. Ultrasonic testing, acoustic emission and eddy current testing method and so on regularly check railway wheelset in service. However, it is difficult to detect a crack initiation clearly with ultrasonic testing due to noise echoes. It is necessary to develop a non-destructive technique that is superior to conventional NDT techniques in order to ensure the safety of railway wheelset. In the present paper, the new NDT technique is applied to the detection of surface defects for railway wheelset. To detect the defects for railway wheelset, the sensor for defect detection is optimized and the tests are carried out with respect to surface and internal defects each other. The results show that the surface crack depth of 1.5 mm in press fitted axle and internal crack in wheel could be detected by using the new method. The ICFPD method is useful to detect the defect that initiated in railway wheelset.

• Journal of the Korean Society for Railway
• /
• v.10 no.1 s.38
• /
• pp.1-6
• /
• 2007

Railway wheels in service are regularly checked by ultrasonic testing, acoustic emission and eddy current testing method and so on. However, ultrasonic testing is sometimes inadequate for sensitively detecting the cracks in railway wheel which is mainly because of the fact of crack closure. Recently, many researchers have actively fried to improve precision for defect detection of railway wheel. The development of a nondestructive measurement tool for wheel defects and its use for the maintenance of railway wheels would be useful to prevent wheel failure. The induced current focusing potential drop(ICFPD) technique is a new non-destructive tasting technique that can detect defects in railway wheels by applying on electro-magnetic field and potential drops variation. In the present paper, the ICFPD technique is applied to the detection of surface and internal defects for railway wheels. To defect the defects for railway wheels, the sensor for ICFPD is optimized and the tests are carried out with respect to 4 surface defects and 6 internal defects each other. The results show that the surface crack depth of 0.5 mm and internal crack depth of 0.7 mm in wheel tread could be detected by using this method. The ICFPB method is useful to detect the defect that initiated in the tread of railway wheels

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.4
• /
• pp.247-252
• /
• 2016

In this paper, we propose a new defect detection technology for micro-defect on the surface of steel products. Due to depth and size of microscopic defect, slop of surface and vibration of strip, the conventional optical method cannot guarantee the detection performance. To solve the above-mentioned problems and increase signal to noise ratio, a novel retro-schlieren method that consists of retro reflector and knife edge is proposed. Moreover dual switching lighting method is also applied to distinguish uneven micro defects and surface noise. In proposed method, defective regions are represented by a black and white pattern. This pattern is detected by a defect detection algorithm with Gabor filter. Experimental results by simulator for sample defects of cold rolled steel show that the proposed method is effective.

• 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.

• Proceedings of the KSR Conference
• /
• 2007.11a
• /
• pp.1940-1945
• /
• 2007

The wheelsets are one of most important component: damages in wheel tread and press fitted axle are a significant cost for railway industry. Since failure in railway wheelset can cause a disaster, regular inspection of defects in wheels and axles are mandatory. Ultrasonic testing, acoustic emission and eddy current testing method and so on regularly check railway wheelset in service. However, it is difficult to use this method because of its high viscosity and because its sensitivity is affected by temperature. Also, due to noise echoes it is difficult to detect defects initiation clearly with ultrasonic testing. It is necessary to develop a non-destructive technique that is superior to conventional NDT techniques in order to ensure the safety of railway wheelset. In the present paper, the new NDT technique is applied to the detection of surface defects for railway wheelset. To detect the defects for railway wheelset, the sensor for defect detection is optimized and the tests are carried out with respect to surface and internal defects each other. The results show that the surface crack depth of 1.5 mm in press fitted axle and internal crack in wheel could be detected by using the new method. The ICFPD method is useful to detect the defect that initiated in the tread of railway wheelset.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.2
• /
• pp.136-146
• /
• 2022

Some surface defects make the weak quality of steel materials. To limit these defects, we advocate a one-stage detector model RetinaNet among diverse detection algorithms in deep learning. There are several backbones in the RetinaNet model. We acknowledged two backbones, which are ResNet50 and VGG19. To validate our model, we compared and analyzed several traditional models, one-stage models like YOLO and SSD models and two-stage models like Faster-RCNN, EDDN, and Xception models, with simulations based on steel individual classes. We also performed the correlation of the time factor between one-stage and two-stage models. Comparative analysis shows that the proposed model achieves excellent results on the dataset of the Northeastern University surface defect detection dataset. We would like to work on different backbones to check the efficiency of the model for real world, increasing the datasets through augmentation and focus on improving our limitation.