• 한국컴퓨터정보학회논문지
• /
• 제14권1호
• /
• pp.99-106
• /
• 2009

경계 검출은 영상 인식의 성능을 좌우하는 중요한 단계로서 지금까지 많은 경계 검출 방법들이 연구되어 왔음에도 불구하고 여전히 어려운 문제로 남아있으며, 영상에서 객체를 인식하거나 특정 영역을 추적하는 등의 유일한 방법은 아니지만 영상 인식 응용 분야에서 중요한 요소들 중의 하나이다. 본 논문에서는 미분 연산자를 이용한 경계검출 방법과는 달리 이진화 영상에서 2개의 이웃된 픽셀 정보를 참조하여 미리 정의된 4개의 경계 픽셀 패턴과 비교하여 경계 픽셀을 찾고, 다음 경계 검출 탐색 픽셀에 대한 방향을 결정하여 영역의 경계를 검출하며, 다음 영역의 경계 검출을 위하여 방문하지 않은 픽셀들을 탐색하면서 경계 검출 단계를 반복 수행하여 이진화 영상의 경계를 검출하는 방법을 제안하였다. 영상 인식에 있어 미분 연산자를 이용하여 경계 검출을 할 경우 경계 검출 다음 단계인 세선화 과정을 생략할 수 있었으며, 기존의 영역 경계 추적 방법에 비해 경계 검출 알고리즘 수행 시간을 단축하여 실시간으로 영상을 인식하는 시스템에 적용하여 전체적인 영상 인식 시간을 단축할 수 있었다.

• Journal of Welding and Joining
• /
• 제24권4호
• /
• pp.32-38
• /
• 2006

The objective of present research works is to investigate the effects of process parameters, including the power of laser, cutting speed, material thickness, and the edge angle, on the melted area in the sharp edge of the cut material fur the case of cutting of a low carbon steel sheet using high-power CW Nd:YAG laser. In order to investigate the influence of edge angle and size of loop on the melted area in the sharp edge, angular cutting tests and loop cutting tests have been carried out. From the results of angular cutting tests, the relationship between the edge angle and the melted area has been obtained. The results of the experiments have been shown that the melted area is rapidly reduced from $120^{\circ}$ of the edge angle and the melted area is nearly zero at $150^{\circ}$ of the edge angle. Through the results of loop cutting experiments, the relationship between the cutting angle on the melted area in the edge according to the size of loop have been obtained. In addition, it has been shown that a proper size of loop is nearly 3 mm as the corner angle is greater than $90^{\circ}$ and 5 mm as the comer angle is less than $90^{\circ}$. The results of above experiments will be reflected on the knowledge base to generate optimal cutting path of the laser.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅳ
• /
• pp.2092-2095
• /
• 2003

In this paper, we propose an algorithm to extract rectangular object area such 3s Data Matrix two-dimensional barcode using edge tracing-based linear feature detection. Hough transform is usually employed to detect lines of edge map. However, it requires parametric image space, and does not find the location of end points of the detected lines. Our algorithm detects end points of the detected lines using edge tracing and extracts object area using its shape information.

• 대한치과기공학회지
• /
• 제18권1호
• /
• pp.117-127
• /
• 1996

This study was aimed to help the construction of esthetic dental prosthesis by investigation of the factors affecting on the atterition position and attrition angle of maxillary lateral incisors. Therefore 197 complete cast of maxillary and mandibualar extracted form the student of K. college were subjected for this study, and result throught the study are as follows. 1. None attrite rate of the maxiilary right lateral incisors was about 16.2% and that of the maxillary left lateral incisors was about 32.4% of examined teeth. 2. Throught mesiodistal attrition area 1) It showed that right lateral incisors was the most frequence in attrition of mesial area of incisal edge, and left lateral incisors was the most frequence in attrition of mesial and mid area of incisal edge. 2) It showed that square type arch was more frequence in attrition of all incisal edge, and ovoid type arch was more frequence in the attrition of mid area of incisal edge, and ovoid type arch was more frequence in the attrition of mid area of incisal edge, and taper type arch was more frequence in the attrition of mesial area of incisal edge than it of average frequence of right lateral incisors, by dental arch type. 3) It showed that square type arch was more frequence in the attrition of all area and mid area of incisal edge, and ovoid type arch was more frequence in the attrition of mid area of incisal edge, and taper type arch was more frequence in the attrition of mesial and distal area of incisal edge than it of average frequence of left lateral incisal, by dental arch type. 4) Sex, vertical overlap, horizontal overlap, incisal guide angle, did not affect significantly to throughout mesiodistal attrition, statistically 3. Throughout labiolingual attrition quantity. 1) It showed that throughout labiolingual attrition quantity was more attrition in order of taper type arch < ovoid type arch < square type arch, by dental arch type. 2) It showed that throughout labiolingual attrition qauntity was more attrition when the length of horizontal overlap is shorter than it of other, by horizontal overlap. 3) Throughout labiolingual attrition quantity of right lateral incisors showed that male was more attrition than it of female. 4) Vertical overlap, incisal guide angle, sex on left lateral incisors did not affect significantly to throughout labiolingual attrition, statistically. 4. Attrition angle 1) It showed that average attrition anlge of right lateral incisors were $30{\pm}13.02$ degree, and it of left lateral incisors were $26{\pm}13.37$ degree. 2) It showed that taper type arch have a bigger attrition angle than it of average of lateral incisors, and square tape arch have a smaller attrition angle than it of average of lateral incisors, by dental arch type. 3) It showed that horizontal overlap of 2.1mm above have a bigger attrition angle than it of average, by horizontal overlap. 4) It showed that female have a bigger attrition angle it of male, by sex.

• 농촌계획
• /
• 제21권4호
• /
• pp.27-33
• /
• 2015

Forest ecosystem is considered as an important stepping stone to minimize the impact of climate change. However, the rapid urbanization has caused fragmentation of forest ecosystem. The fragmentation of forest patch results in edge effect which brings about adverse impacts on forest function and structure. Degradation of forest ecosystem decreases carbon sequestration because edge effect reduces productivity. Therefore, we analyzed the impact of forest edge effect on forest ecosystem carbon stock change in Seongnam-si, Gyeonggi-do. We used connectivity analysis to determine forest edge and core area. The field study sites were selected with considering forest age, density, class and soil type. Secondly, forest carbon stock was calculated with allometric equation. The soil carbon stock was derived from Walkely-Black method. Lastly, Mann-Whitney test was conducted to validate differences between carbon stock in edge and core area. As a result of study, the connectivity analysis was effective to determine forest edge and core. The core and edge of forest patch showed different composition of tree species and soil properties. Carbon stock per tree in the edge area was lower than that in the core area. However, the difference of soil organic carbon content between the edge and core were relatively small. This assessment can be applied for the conservation of forest patch as well as quantitative assessment on the forest carbon stock change caused by fragmentation.

• 대한임베디드공학회논문지
• /
• 제17권2호
• /
• pp.77-83
• /
• 2022

In this paper, the two-stage object detection approach is proposed to implement a deep learning-based label inspection system on edge computing environments. Since the label printed on the products during the production process contains important information related to the product, it is significantly to check the label information is correct. The proposed system uses the lightweight deep learning model that able to employ in the low-performance edge computing devices, and the two-stage object detection approach is applied to compensate for the low accuracy relatively. The proposed Two-Stage object detection approach consists of two object detection networks, Label Area Detection Network and Character Detection Network. Label Area Detection Network finds the label area in the product image, and Character Detection Network detects the words in the label area. Using this approach, we can detect characters precise even with a lightweight deep learning models. The SF-YOLO model applied in the proposed system is the YOLO-based lightweight object detection network designed for edge computing devices. This model showed up to 2 times faster processing time and a considerable improvement in accuracy, compared to other YOLO-based lightweight models such as YOLOv3-tiny and YOLOv4-tiny. Also since the amount of computation is low, it can be easily applied in edge computing environments.

• 응용통계연구
• /
• 제25권6호
• /
• pp.1019-1026
• /
• 2012

This paper presents a method for the identification of "edge observations" located on a boundary area constructed by a truncation variable as well as for the identification of outliers and the after fit of multivariate skew $t$-distribution(MST) to asymmetric data. The detection of edge observation is important in data analysis because it provides information on a certain critical area in observation space. The proposed method is applied to an Australian Institute of Sport(AIS) dataset that is well known for asymmetry in data space.

• 한국컴퓨터정보학회논문지
• /
• 제22권7호
• /
• pp.47-54
• /
• 2017

Recently, correlation filter based trackers have shown excellent tracking performance and computational efficiency. In order to enhance tracking performance in the correlation filter based tracker, search area which is image patch for finding target must include target. In this paper, two methods to discriminatively represent target in the search area are proposed. Firstly, search area location is estimated using pyramidal Lucas-Kanade algorithm. By estimating search area location before filtering, fast motion target can be included in the search area. Secondly, we investigate multi-channel Local Edge Pattern(LEP) which is insensitive to illumination and noise variation. Qualitative and quantitative experiments are performed with eight dataset, which includes ground truth. In comparison with method without search area estimation, our approach retain tracking for the fast motion target. Additionally, the proposed multi-channel LEP improves discriminative performance compare to existing features.

• 한국복합재료학회:학술대회논문집
• /
• 한국복합재료학회 2001년도 춘계학술발표대회 논문집
• /
• pp.78-82
• /
• 2001

Impact behaviors of Aluminum Honeycombs Sandwich Panel(AHSP) by drop weight test were investigated. Two types of specimens with 1/2" and 1/4" cell size were tested by two impactors which are weight of $5.25\textrm{kg}_{\textrm{f}}$ and $11.9\textrm{kg}_{\textrm{f}}$. Parametric studies were achieved including the impactor weight and impact sites which consist face, long-edge, short-edge, and point of the specimen. Face one of impact sites was the strongest and short-edge one of impact sites was the weakest. The damaged area of AHSP was enlarged with the increase of impactor weight that is equal to impact energy. After 3 point bending test, fracture modes of AHSP were analyzed with AE counts. Lower facesheet was fractured in the long-edge direction and then separated between facesheet and core. In the short-edge direction after core wrinkled, lower facesheet tear occurred. Impact behavior by FE analysis were increased localized damage in fast velocity because the faster velocity of the impact was, the smaller the stress of core was. Consequently, impactor weight had an effect on widely damaged area, while the impact velocity was caused on the localized damaged area.aged area.

• 한국통신학회논문지
• /
• 제21권10호
• /
• pp.2541-2551
• /
• 1996

인간의 시각체계는 영상의 밝기의 정도가 균일한 연에서는 잡음에 민감하지만, 변화하는 부분에서는 에지(edge)의 정도가 심할수록 잡음에 둔감하고, 에지부분에서 멀어질수록 잡음에 대한 민감도가 급격하게 증가한다. 이러한 인간의 시각 특성에 기반을 둔 여러가지 영상복원 방식이 제안되고 있는데, 본 논문에서는 영상을 복원함에 있어서 윤곽 부분에서는 변화하는 부분의 선명도를 높이고, 영상이 평탄한 부분에서는 잡음 성분을 많이 억제 시켜서 영상을 주관적으로 향상시키는 적응적 영상복원 방식을 소개한다. 이 방법은 에지 검출을 하기 위해서 각 화소를 기준으로 지역 분산값(local variance)을 사용하여 시각 함수(visibility function)를 구하고, 이 값에 따라 정규화 매개변수를 변환시켜 적응적으로 영상을 복원한다. 즉 영상을 평탄한 부분에서 에지부분까지 몇 단계로 나누어서 각각의 단계에 해당하는 유한 임펄스 CLS 필터를 구현해서 영상을 복원한다.