http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
하종은 제어·로봇·시스템학회 2012 International Journal of Control, Automation, and Vol.10 No.6
Sensor fusion of a camera and laser range finder is important for the autonomous navigation of mobile robots. Finding the transformation between the camera and laser range finder is the first nec-essary step for the fusion of information. Many algorithms have been proposed, but these tend to re-quire many different steps in order to achieve reliable and accurate results. A calibration structure that has triangular hole on its plane is proposed for the extrinsic calibration of a camera and laser range finder. Locations of laser scan data that are invisible on the calibration plane can be determined using property on the proposed calibration structure. First, we classify the laser scan data into two groups where one is on the plane and the other is off the plane. Then, we determine the absolute location of the laser scan data on the plane through a search of the parameters of the line. Finally, we can establish 3D-3D correspondences between the camera and laser range finder. Extrinsic calibration between a camera and laser range finder is found using a conventional 3D-3D transformation computing algorithm.
An Image Processing Algorithm for the Automatic Manipulation of Tie Rod
하종은 제어·로봇·시스템학회 2013 International Journal of Control, Automation, and Vol.11 No.5
In this paper, we present a robust image processing algorithm for the automatic manipulation of tie rod in the automotive assembly process. A sensing system consisting of a camera and a slit laser is used to obtain 3D data for processing. It is attached to the robot’s arm. A nut runner is used as an end-effector of the robot to manipulate the bolt and nut on the tie rod. For the successful operation, a robust image processing algorithm that can cope with diverse illumination variations in manufacturing conditions is required. We present a robust algorithm for the extraction of laser contours on images. It consists of two steps where color and grey information is used to cope with clutters caused by laser reflections on metal surfaces. Experimental results using many images under real conditions show the robustness of the proposed algorithm.
하종은 제어·로봇·시스템학회 2015 International Journal of Control, Automation, and Vol.13 No.5
Extrinsic calibration between a camera and laser range finder (LRF) is required to integrate information from both sensor. In this paper, we propose an improved algorithm of our earlier work [1] for the extrinsic calibration of a camera and laser range finder. We improve our previous algorithm in two directions. First, control points on laser range finder’s data are obtained by finding the crossing point of two fitted lines where previous algorithm directly used laser range finder’s data. It can reduce the effect of noisy measurement of laser range finder. Second, difference of distance on calibration plane is used as cost where previous algorithm used matching portion. Proposed algorithm gives an improved result than previous one. It is verified thorough experimental results.
내시경적으로 제거한 식도의 과립 세포 종양 -증례보고 및 국내 문헌 고찰-
하종,이옥재,조활석,정태식,윤지향,이은정,민현주,김태효,정운태,조중현 대한소화기내시경학회 2003 Clinical Endoscopy Vol.26 No.2
Granular cell tumor is a benign tumor, commonly found in the skin, tongue, and breast but rarely in the esophagus. A 44-year-old man was referred from the private clinic because of an esophageal lesion found on endoscopy. Esophagoscopy revealed a white-yellowish polypoid lesion covered with normal looking mucosa in the upper esophagus. It was movable within the wall by pushing with forceps. Endoscopic polypectomy after ligation with elastic "O" band was performed without complication. Histologic diagnosis of granular cell tumor was made. It was stained strongly positive for S-100 protein. Another 20 cases of esophageal granular cell tumors reported in Korean literature were reviewed. (Korean J Gastrointest Endosc 2003;26:84-89) 과립세포종은 양성 종양으로 피부, 구강, 유방 등에서는 흔하지만, 식도에서는 드물게 발생한다. 저자들은 식후 상복부 불편감으로 위내시경 검사를 받은 후 식도의 점막하종양을 발견하고 내원한 44세 남자에서, ‘O’형 밴드 결찰술을 이용한 용종절제술을 시행하고 조직학적으로 과립세포종으로 확진하였기에 국내의 문헌고찰을 중심으로 보고한다.
합성 이미지를 이용한 Mask R-CNN 기반 한국 번호판 검출
하종은 제어·로봇·시스템학회 2020 제어·로봇·시스템학회 논문지 Vol.26 No.9
License plate detection on an image is the first necessary step for the automatic recognition of license plate. In this paper, we adopt Mask R-CNN [11] to detect a license plate on an image. It requires many training images to cope with over-fitting that occurs when training samples are smaller than numbers of parameters. In general, more than 200K images are required for the stable training, but it requires large amount of time and cost. In this paper, we present a method that uses already available open dataset. We use two open dataset of CCPD [8] and BDD [18]. CCDP dataset provides locations of four corner points on an image. But, they contain license plate of China. First, Korean license plate image is made by referencing the design rule. Then, Korea license plate image is projected onto the corresponding positions of CCPD image. Four corresponding points between Korea license plate and CCPD images is used in the computation of perspective transform. Two different types of training images from CCPD and BDD dataset are used in the training of Mask R-CNN, and they are applied to image that contains real Korea license plate. Training using composite images from CCPD shows better performance than that of BDD on the real Korea license plate image. Experimental results show the feasibility of presented approach. .
Grouping Contents on Korean Road Signs
하종은 제어·로봇·시스템학회 2011 International Journal of Control, Automation, and Vol.9 No.6
In this paper, we propose a method for grouping the content of Korean road signs into the five categories of Korean, English, direction symbol, road symbol and distance digit. The first step necessary to automate the inspection of road signs is to determine whether their content is reproduced as clearly as in the original design. Previously, inspections were performed manually by humans. In-stead, we deal with images that are acquired manually by digital cameras, using various features in-cluding color, relative length and size, reflecting the design rule for Korean road sign. We began by analyzing blobs that are obtained through connected component analysis after binarization. These are first grouped into two categories: direction symbol and size. Then, the road symbol is selected using color. Finally, the remaining blobs are grouped using the relative length between blobs. Experimental results using real images show the feasibility of the proposed algorithm.
Automatic Registration of Two Parts using Robot with Multiple 3D Sensor Systems
하종은 대한전기학회 2015 Journal of Electrical Engineering & Technology Vol.10 No.4
In this paper, we propose an algorithm for the automatic registration of two rigid parts using multiple 3D sensor systems on a robot. Four sets of structured laser stripe system consisted of a camera and a visible laser stripe is used for the acquisition of 3D information. Detailed procedures including extrinsic calibration among four 3D sensor systems and hand/eye calibration of 3D sensing system on robot arm are presented. We find a best pose using search-based pose estimation algorithm where cost function is proposed by reflecting geometric constraints between sensor systems and target objects. A pose with minimum gap and height difference is found by greedy search. Experimental result using demo system shows the robustness and feasibility of the proposed algorithm.