• Journal of Intelligence and Information Systems
• /
• v.2 no.1
• /
• pp.1-9
• /
• 1996

사출 성형의 초보자도 이용할 수 있으며 사출 성형기의 이력 관리를 할 수 있는, 플라스틱 사출 성형의 진단과 불량대책을 위한 지식 기반 전문가 시스템을 ATR-IM 이라는 범용 전문가시스템 쉘을 이용하여 구현하였다. 지식베이스(knowledge base)는 수지 회사와 관련 서적, 전문가의 경험을 토대로 구축하였으며, 구현된 시스템은 크게 사출 공정 조건 설정 시스템, 사출기 ID에 따른 이력 관리 시스템, 사출성형 불량진단 시스템의 세부분으로 나누어 구성하였다. 본 연구를 통해 제안된 방법에 따라 업체별 사출기 성능에서 오는 공정 조건값의 차이를 사출공정 조건값의 저장에 의해 고려할 수 있으며, 불량 해결에 대한 전문가시스템이 추론 순서 결정을 경험적 확률에 따라 이루어지도록 함으로써 현장에서의 작업 환경을 고려한 전문가시스템이 되도록 하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.132-135
• /
• 2021

국내에서 유통되는 일회용 가스라이터 점유율의 약 절반은 국내 유일의 한 공장에서 생산하고 있다. 저렴한 외국산 가스라이터로부터 국내 사업을 보호하기 위해 품질 향상과 원가경쟁력 확보의 중요성이 매우 커진 것이 현실이다. 본 논문에서는 YOLOv4 머신러닝 객체인식 모델과 OpenCV 실시간 이미지 처리 오픈소스를 활용해 개발한 불량품 자동 검출 시스템을 제안한다. 대표적인 불량인 '액화가스 부피 불량품'을 검출하는 시스템을 개발하고 실험을 통해 그 정확성을 검증하였다. 제안한 시스템은 97%의 정확도로 상태를 분류하였으며, 이를 통해 100%의 불량을 검출할 수 있었다.

• Journal of Digital Convergence
• /
• v.12 no.8
• /
• pp.345-350
• /
• 2014

The most important decision of process planning for the manufacturing system is the machine selection problem to minimize machining costs. Each machine has its own different machining performance indicating a different fraction of scrap, making the cost of scrap generated by machining is different for each machine. Therefore, when we decide on machine selection, we must consider the machining cost and the cost of scrap generated. This paper describes the statistical model for the fraction of scrap generated by machining and the machine selection algorithm considering the total cost including the machining cost and the cost of scrap generated.

• Journal of Korea Society of Industrial Information Systems
• /
• v.18 no.5
• /
• pp.45-50
• /
• 2013

Recently, as the market for SMART TV expands, the camera is embedded for providing various user experience. However, this leads to occurrence of camera failure due to TV power up sequence problem, which are usually not detectable in conventional test equipments. Although the failure-detection can be possible by re-generating control signals for audio interface with new equipment, it is expensive and also requires much time to test. In this paper, for SMART TV, FPGA(Field Programmable Gate Array)-based failure-detection system is proposed which can lead to reduction of both cost and time for test.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.43 no.9 s.351
• /
• pp.38-43
• /
• 2006

In this paper, we propose the efficient dead pixel detection algorithm for CMOS image sensors and its hardware architecture. The CMOS image sensors as image input devices are becoming popular due to the demand for miniaturized, low-power and cost-effective imaging systems. However, the presence of the dead pixels degrade the image quality. To detect the dead pixels, the proposed algorithm is composed of scan, trace and detection step. The experimental results showed that it could detect 99.99% of dead pixels. It was designed in a hardware description language and total logic gate count is 3.2k using 0.25 CMOS standard cell library.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.29 no.6
• /
• pp.1185-1189
• /
• 2000

본 연구는 입원한 위암환자를 대상으로 영양불량의 정도를 중심으로 영양위험의 정도를 추정하며, 이러한 환자들의 초기 영양상태와 치료결과와의 관련성을 규명하기 위해 시도되었다. 연구대상은 병원에 입원한 209명의 위암환자였다. 영양불량은 표준체중백분율, 혈청알부민, 총임파구(TLC)로 판정하였으며, 치료결과는 합병증유무, 퇴원시 상태, 사망 등 세가지 관점에서 측정하였다. 표준체중백분율, 혈청 알부민, 총임파구수를 기준으로 영양상태를 보았을 때 비위험군은 39.7%였고, 위험군 I은 41.6%, 영양불량이 심한 위험군 II는 18.7%로, 환자의 60.3%가 영양상태가 불량한 것으로 나타나 위암환자 입원초기의 영양불량 정도가 심한 것으로 나타났다. 치료 결괴 퇴원시 상태가 "좋은"이 95.7%, 나쁨이 "4.3%"였으며, 사망환자는 6.8%, 합병증은 20.1%발생하였다. 또한 초기영양상태와 합병증과의 관련성은 높은 변수군 분류의 적절성(p<0.03)을 보여주고 있으며, 초기영양상태와 퇴원시의 치료상태에서도 높은 변수군 분류의 적절성(p<0.001)이 있는 것으로 나타났다. 환자의 초기 영양상태와 사망과의 관련성 검토 결과 초기영양상태가 불량한 경우 사망확률이 높은 것으로 나타나높은 상관성이 있음을 보였다. 이상의 연구결과로 볼 때 위암 환자의 초기 영양상태는 환자의 치료결과와 높은 관련성이 있는 것으로 나타났다.높은 관련성이 있는 것으로 나타났다.

• The KIPS Transactions:PartB
• /
• v.17B no.6
• /
• pp.445-458
• /
• 2010

A Cold Cathode Fluorescent Lamp(CCFL) is used as a LCD Monitor's backlight widely. The most common way to check CCFL's defects is an examination with the naked eye. This naked eye examination can cause examination inconsistencies and industrial disasters. A shooting environment and detection algorithms are important for finding CCFL defects automatically. This paper presents CCFL defect detection algorithms using images captured under the shooting environment with sidelight which is one of the shooting environment we have suggested. The experimental result shows 4.65% of overdetection and 5.37% of unsuccessful defect detection of CCFL.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.5
• /
• pp.1053-1059
• /
• 2012

Barcode labels are being widely used for identifying products since they are cheap and easy to use. As the barcode labels are massively produced by seal printing, some labels have defects such as poor printing quality or data mismatch between barcode and the text. Barcode read errors and business errors caused by defected barcodes result in delay in logistics and increased processing costs. In this paper, we propose an on-line barcode verification system that uses image processing technique to verify the quality of labels at the production stage. The proposed system captures label images through the vision camera and then checks the print quality and verifies the combination of barcodes and texts in a label. If any defected label is found, the proposed system gives alarm signals and marks the defected labels so that they are removed at early stage of the production.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.7
• /
• pp.1551-1556
• /
• 2010

In this paper, we propose a new inspection method that applies fuzzy reasoning method considering the difference of brightness and intensity of illumination by bend together. In the preprocessing phase, we compensate the degree of semiconductor images with bilinear interpolation and moment-rotation. Then we use fuzzy reasoning method with the difference of brightness from error region by pattern matching and the difference of intensity of illumination from bends. Then the result is difuzzified and applied to the final inspection process. In experiment which uses 30 real world semiconductors with strait shots and side shots, the proposed method successfully discard the false positive identified by conventional brightness comparison only method without any loss of misidentification.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.47 no.12
• /
• pp.24-30
• /
• 2010

In recent memories, repair is an unavoidable method to maintain its yield and quality. The probability of defect occurence on spare lines has been increased through the growth of the density of recent memories with 2 dimensional spare architecture. In this paper, a new analysis region virtualization scheme is proposed. the analysis region virtualization scheme can be applied with any BIRA (built-in redundancy analysis) algorithms without the loss of their repair rates. The analysis region virtualization scheme can be a viable solution for BIRA considering the faulty spare lines of the future high density memories.