• 한국컴퓨터산업학회논문지
• /
• 제3권12호
• /
• pp.1725-1736
• /
• 2002

에지 시그널의 세기가 갖는 특징적인 성질로서, 에지를 가로질러 엄격하게 단조적인 세기 변화가 나타남을 설명하고 이를 바탕으로 하는 새로운 에지 검출 알고리듬을 제안한다. 스케일링에 무관하고 비국소적인 확장된 방향미분을 픽셀 공간에서 도입하여, 본 알고리듬이 에지 폭의 다양한 변화에 적응성을 가지며 최적의 에지 검출 알고리듬으로서 적절함을 설명한다. 본 알고리듬의 산업적 응용으로서, 시각적 검사에 대한 간단한 컴퓨터 비전 프로시저의 한 예를 살펴본다.

• 센서학회지
• /
• 제33권3호
• /
• pp.179-184
• /
• 2024

In the car industry, welding is a fundamental linking technique used for joining components, such as steel, molds, and automobile parts. However, accurate inspection is required to test the reliability of the welding components. In this study, we investigate the detection of weld beads using 2D image processing in an automatic recognition system. The sample image is obtained using a 2D vision camera embedded in a lighting system, from where a portion of the bead is successfully extracted after image processing. In this process, the soot removal algorithm plays an important role in accurate weld bead detection, and adopts adaptive local gamma correction and gray color coordinates. Using this automatic recognition system, geometric parameters of the weld bead, such as its length, width, angle, and defect size can also be defined. Finally, on comparing the obtained data with the industrial standards, we can determine whether the weld bead is at an acceptable level or not.

• 전기학회논문지P
• /
• 제52권1호
• /
• pp.35-41
• /
• 2003

In this paper, a machine vision system for recognizing and classifying the patterns and marks engraved by die molding or laser marking on the glass panels of computer monitors is suggested and evaluated experimentally. The vision system is equipped with a neural network and an NGC pattern classifier including searching process based on normalized grayscale correlation and adaptive binarization. This system is found to be applicable even to the cases in which the segmentation of the pattern area from the background using ordinary blob coloring technique is quite difficult. The inspection process is accomplished by the use of the NGC hypothesis and ANN verification. The proposed pattern recognition system is composed of three parts: NGC matching process and the preprocessing unit for acquiring the best quality of binary image data, a neural network-based recognition algorithm, and the learning algorithm for the neural network. Another contribution of this paper is the method of generating the training patterns from only a few typical product samples in place of real images of all types of good products.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2009년도 정보 및 제어 심포지움 논문집
• /
• pp.91-93
• /
• 2009

In this paper, we present the Q-learning method for adaptive traffic signal control on the basis of In this paper, we present the Q-learning method for adaptive traffic signal control on the basis of multi-agent technology. The structure is composed of sixphase agents and one intersection agent. Wireless communication network provides the possibility of the cooperation of agents. As one kind of reinforcement learning, Q-learning is adopted as the algorithm of the control mechanism, which can acquire optical control strategies from delayed reward; furthermore, we adopt dynamic learning method instead of static method, which is more practical. Simulation result indicates that it is more effective than traditional signal system.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2007년도 하계종합학술대회 논문집
• /
• pp.361-362
• /
• 2007

We propose a contrast-controlled feature detection approach for steel radiograph image. X-ray images are low contrast, dark and high noise image. So, It is not simple to detect defects directly in automated radiography inspection system. Contrast enhancement, histogram equalization and median filter are the most frequently used techniques to enhance the X-ray images. In this paper, the adaptive control method based on contrast limited histogram equalization is compared with several histogram techniques. Through comparative analysis, CLAHE(contrast controlled adaptive histogram equalization) can enhance detection of defects better.

• 한국항공운항학회지
• /
• 제11권2호
• /
• pp.33-42
• /
• 2003

In this paper, we propose an enhancement of the ultrasonic image for non-destructive inspection of aircraft composite materials. The ultrasonic images are corrupted by a speckle noise which has the characteristic of granular pattern noise and is in all types of coherent imaging systems, the signal independent and multiplicative noise. In this paper, we derive a filter, called the AWLF(Adaptive Window Log Filter), from the nature of the speckle. The filter is made of the MEAN Filter in the edge region and Log Filter in the flat or noise region. To make a distinction between edge and flat region, we calculate the inclination around the local window instead of computing the local statistics of pixels such as local mean ${\bar{M}}$ and standard deviation ${\sigma}_s$. According to the obtained region, edge region is performed by the mean filter and flat region by the Log filter. Performance of the proposed filter is evaluated by the Enhanced Factor$(F_e)$ and the Speckle Index(SI).

• 정보처리학회논문지B
• /
• 제9B권2호
• /
• pp.245-254
• /
• 2002

머신비전을 이용한 IC 패키지 마킹검사 시스템은 입력영상으로부터 검사할 요소들의 위치를 식별하고, 추출된 요소들을 학습된 표준 패턴과 비교하여 마킹의 불량 여부를 판단한다. 본 논문에서는 검사 대상 IC 패키지의 위치 판별, 마킹문자 추출, 핀원딤플 검출과 같은 일련의 작업들에 적합한 적응적 다단계 이진화 방법과 마킹문자의 국소적인 오류검출은 물론 잡영에 강건한 정합단위의 동적 선택 방법을 제안한다. 제안하는 이진화 방법은 이진화 대상 영역과 명도 값의 범위를 제한하여 Otsu의 이진화 알고리즘을 적용함으로써 특정 응용에 적응적인 이진화가 가능하다. 정합단위의 동적 선택 방법은 문자추출 및 배치분석에 대한 결과에 따라 정합단위를 선택한다. 그러므로 문자추출 및 배치분석 과정에서 발생하는 예기치 못한 부적절한 상황에서도 가능한 범위내에서 최소의 정합단위를 선택할 수 있다. 제안된 방법을 구현하여 8종의 IC 패키지, 총 280개의 영상에 대하여 실험한 결과, IC 패키지와 핀원딤플의 검출율은 100%였으며, 마킹상태에 대한 판정은 98.8%의 정확도를 나타내어 제안된 방법이 효과적임을 확인할 수 있었다.

• 한국콘텐츠학회논문지
• /
• 제11권3호
• /
• pp.18-25
• /
• 2011

현재 주화의 제조 공정에서는 주화의 표면 품질 진단을 사람이 눈으로 직접 확인하여 수행하고 있다. 본 논문은 컨베이어 벨트에 놓이어 이동하는 주화로부터 획득한 영상을 이용하여 주화 표면의 결함을 검출하는 영상처리 방법을 제시한다. 결함 검출 방법은 영상에서 주화 영역을 분할하고, 분할된 동전을 비교할 모델에 정렬하며, 정렬된 영상을 최적의 고유 영상 공간으로 투영, 투영 오차와 학습된 가변 임계값과 비교하여 결함 부위를 검출한다. 본 논문에서는 이러한 일련의 영상처리 과정 중에서 주화 표면 진단과 관련하여 특화된 새로운 방법을 제시한다. 주화의 정렬을 위하여 분할된 주화의 히스토그램을 사용한다. 이 방법은 2차원 영상의 정렬을 일차원 히스토그램의 정렬로 변환하는 것이다. 다음으로 정렬된 영상을 고유 영상공간에 투영시켜 주화 방향에 따른 휘도 변화를 보정한다. 이 방법은 소수의 고유 영상 벡터들로 구성된 고유 영상 공간을 여러 개 생성하고, 최적의 고유 영상 공간에 정렬된 영상을 투영하여 실시간 구현이 가능하게 한다.

• 한국정밀공학회지
• /
• 제22권12호
• /
• pp.51-60
• /
• 2005

This paper describes a robust control scheme for high-speed and long stroke scanning motion of high precision linear motor system consisting of linear motor, air bearing guide and position measurement system using heterodyne interferometer. Nowadays, semiconductor process and inspection of wafer or LCD need high speed and long travel length for their high throughput and extremely small velocity fluctuations or tracking errors. In order to satisfy these conditions, linear motor system are widely used because they have large thrust force and do not need motion conversion mechanisms such as ball screw, rack & pinion or capstan with which the system are burdened. However linear motors have a problem called force ripple. Force ripple deteriorates the tracking performances and makes periodic position errors. So, force ripple must be compensated. To maximize the tracking performance of linear motor system, we propose the control scheme which is composed of a robust control method, Time Delay Controller (TDC) and a feedforward control method, Zero Phase Error Tracking Control (ZPETC) for accurate tracking a given trajectory and an adaptive force ripple compensation (AFC) algorithm fur estimating and compensating force ripple. The adaptive ripple compensation is continuously refined on the basis of tracking error. Computer simulation results based on modeled parameters verify the effectiveness of the proposed control scheme for high-speed, long stroke and high precision scanning motion and show that the proposed control scheme can achieve a sup error tracking performance in comparison to conventional TDC control.

• Journal of Advanced Marine Engineering and Technology
• /
• 제39권2호
• /
• pp.179-186
• /
• 2015

In the indoor location estimation system, which has recently been actively studied, the received signal strength indicator contains a high level of noise when measuring the signal strength in the range between two nodes consisting of a receiver and a transceiver. To minimize the noise level, this paper proposes an improved adaptive smoothing filter that provides different exponential weights to the current value and previous averaged one of the data that were obtained from the nodes, because the characteristic signal attenuation of the received signal strength indicator generally has a log distribution. The proposed method can effectively decrease the noise level by using a feedback filter that can provide different weights according to the noise level of the obtained data and thus increase the accuracy in the distance and location without an additional filter such as the link quality indicator, which can verify the communication quality state to decrease the range errors in the indoor location recognition using ZigBee based on IEEE 802.15.4. For verifying the performance of the proposed improved adaptive smoothing filter, actual experiments are conducted in three indoor locations of different spatial sections. From the experimental results, it is verified that the proposed technique is superior to other techniques in range measurement.