에지방향 및 수평 움직임을 이용한 시-공간 디인터레이싱 기법

김동형(Dong-Hyung Kim) 한국정보기술학회 2015 한국정보기술학회논문지 Vol.13 No.3

Interlaced video, in which one frame consists of even and odd fields that have 1/60 sec time interval, yields some artifacts like jitter effect owing to those time interval when displaying one frame. Deinterlacing to solve this problem divides into spatial and temporal methods. Each method has strong and weak points in terms of hardware implementation, algorithm complexity and performance. In order to get each strong points, the proposed spatio-temporal deinterlacing method using edge direction and horizontal motion defines 9 and 7 cost functions for spatial and temporal deinterlacing, respectively. These functions correspond to different edge directions and horizontal motions. The costs were normalized to compare each other and we can select the best deinterlacing via cost comparison. Simulation results show that the proposed method yields higher objective image qualities than other conventional approaches. Furthermore, subjective image quality of the proposed method also is higher than the others.

Adaptive RRT 를 사용한 고 자유도 다물체 로봇 시스템의 효율적인 경로계획

김동형(Dong-Hyung Kim),최윤성(Youn-Sung Choi),염서군(Rui-Jun Yan),라로평(Lu-Ping Luo),이지영(Ji Yeong Lee),한창수(Chang-Soo Han) 제어로봇시스템학회 2015 제어·로봇·시스템학회 논문지 Vol.21 No.3

This paper proposes an adaptive RRT (Rapidly-exploring Random Tree) for path planning of high DOF multibody robotic system. For an efficient path planning in high-dimensional configuration space, the proposed algorithm adaptively selects the robot bodies depending on the complexity of path planning. Then, the RRT grows only using the DOFs corresponding with the selected bodies. Since the RRT is extended in the configuration space with adaptive dimensionality, the RRT can grow in the lower dimensional configuration space. Thus the adaptive RRT method executes a faster path planning and smaller DOF for a robot. We implement our algorithm for path planning of 19 DOF robot, AMIRO. The results from our simulations show that the adaptive RRTbased path planner is more efficient than the basic RRT-based path planner.

로봇형 차량의 자율주행을 위한 센서 기반 운동 계획법 개발

김동형(Dong-Hyung Kim),김창준(Chang-Jun Kim),이지영(Ji-Yeong Lee),한창수(Chang-Soo Han) 제어로봇시스템학회 2011 제어·로봇·시스템학회 논문지 Vol.17 No.6

This paper presents the motion planning of robotic vehicles for the path tracking and the obstacle avoidance. To follow the given path, the vehicle moves through the turning radius obtained through the pure pursuit method, which is a geometric path tracking method. In this paper, we assume that the vehicle is equipped with a 2D laser scanner, allowing it to avoid obstacles within its sensing range. The turning radius for avoiding the obstacle, which is inversely proportional to the virtual force, is then calculated. Therefore, these two kinds of the turning radius are used to generate the steering angle for the front wheel of the vehicle. And the vehicle reduces the velocity when it meets the obstacle or the large steering angle using the potentials of obstacle points and the steering angle. Thus the motion planning of the vehicle is done by planning the steering angle for the front wheels and the velocity. Finally, the performance of the proposed method is tested through simulation.

MPEG2에서 H.264로의 트랜스코딩 기법

김동형,정제창,Kim, Dong-Hyung,Jeong, Je-Chang 한국통신학회 2005 韓國通信學會論文誌 Vol.30 No.7C

In this paper, we present a transcoding algorithm for converting an MPEG-2 video bitstream to an H.264 bitstream. The proposed transcoder consists of two parts. One is MPEG2 decoding part and the other is H.264 encoding part Because our algorithm is for transcoding in the spatial domain, MPEG2 decoding part carries out full decoding of MPEG2 bitstream. While, because macroblock type and coded block pattern in MPEG2 are significantly related to macroblock mode in H.264, macroblock mode is selected adaptively according to macroblock type and coded block pattern in H.264 decoding part. Furthermore, motion vector is also used as side-information for 16$\ctimes$16 macroblock mode. Simulation results show that the proposed transcoder yields high reduction of total transcoding time at comparable PSNR. 본 논문에서는 MPEG-2로 부호화된 비디오 스트림에서 H.264 부호화 표준으로의 트랜스코딩 기법을 제안한다. 제안하는 MPEG2에서 H.264t로의 트랜스코더는 MPEG2 복호화부와 H.264 부호화부로 구성되어 있으며, MPEG2 복호화부가 입력되는 MEPG2 비디오 스트림의 전복호화(全復號化)를 수행하는 공간영역에서의 트랜스코딩 기법이다. 반면 MPEG2 복호화 과정에서 사용되는 매크로블록 타입과 부호화블록패턴은 H.264의 매크로블록 모드와 상당한 연관성을 가지는데, 이를 이용하여 H.264 부호화부에서는 MEPG2의 매크로블록 타입, 부호화블록패턴 정보를 이용하여 H.264 부호화부에서의 매크로블록 모드를 적응적으로 선택하고 추가적으론 움직임 벡터를 활용함으로 전체 트랜스코더의 복잡도를 감소시킨다. 제안하는 트랜스코더를 시퀀스에 적용한 결과 전복호화 후 전부호화의 결과와 비교하여 평균 0.03dB의 PSNR 차이만을 보이는 반면 전체 트랜스코딩 시간은 최대 61$\%$ 감소된 것으로 나타난다.

조대술 후 구강내 접근술을 통한 거대 법랑모세포종의 치험례 : 증례보고

김동형,조성웅,서동원,강지연,심재환,이동근,김상중,Kim, Dong-Hyung,Cho, Sung-Woong,Seo, Dong-Won,Kang, Ji-Yeon,Sim, Jae-Hwan,Lee, Dong-Keun,Kim, Sang-Jung 대한구강악안면외과학회 2008 대한구강악안면외과학회지 Vol.34 No.2

Ameloblastoma is the most frequently accuring odontogenic tumor in mouth. The biologic behavior of this neoplasm is locally invasive tumor with a high rate of recurrence. But in case of unicystic ameloblastoma, it was known that the neoplasm can be treated by marsupialization and the recurrence rate is lower. In our clinic, we tried to treat one of ameloblastoma cases by marsupialization and finished the treatment by enucleation via intra-oral approach with sagittal ramus osteotomy. This is a report of that case about 29 years old female patient.

H.264/AVC비디오의 고속 부호화를 위한 인트라모드 선택 알고리듬

김동형,정제창,Kim, Dong-Hyung,Jeong, Je-Chang 한국통신학회 2007 韓國通信學會論文誌 Vol.32 No.6c

H.264 부호화 표준은 부호화 효율을 높이기 위해 1/4 화소 단위의 움직임 추정, 다중 참조 프레임, 인트라 예측, 루프 필터, 다양한 블록 크기의 지원 등과 같은 새로운 부호화 도구들을 사용한다. 이를 통해 이전의 비디오 부호화 표준들에 비해 율-왜곡(率-歪曲) 관점에서 높은 성능을 보이지만 그로 인해 부호기의 복잡도는 상당히 증가한다. 본 논문은 부호기 복잡도의 증가를 초래하는 주요 부호화 도구들 중 인트라 매크로블록 모드 선택의 복잡도 감소에 주안(主眼)점을 두며, 이를 위한 고속 부호화 알고리듬을 제시한다. 제안하는 알고리듬은 먼저 간단한 전처리과정을 통해서 Intra4x4 모드의 예측모드를 한정하고, 선택된 Intra4x4의 예측모드를 사용하여 다른 인트라 모드들의 예측모드를 결정한다. 실험결과 제안하는 알고리듬은 기존의 방법보다 높은 성능을 보이며 참조소프트웨어와 비교하여 약 82%의 속도향상을 가져오는 것으로 나타난다. For the improvement of coding efficiency, the H.264 standard uses new coding tools such as VBS, 1/4-pel accurate ME, multiple references, intra prediction, loop filter, etc. Using these coding tools, H.264 has achieved significant improvements from rate-distortion point of view compared to existing standards. However, the encoder complexity is greatly increased due to these coding tools. We focus on the complexity reduction method of intra-mode decision. Our algorithm first restricts selective prediction modes of Intra4x4 using a simple preprocessing. The prediction modes of Intra4x4 are used for restricting those of the other inter-modes. Simulation results show that the proposed method outperforms other conventional methods and save about 82% of total encoding time.

최적 타이어 힘 분배 방법을 통한 전기차의 독립 6WD/6WS에 관한 연구

김동형(Dong-Hyung Kim),김창준(Chang-Jun Kim),김영렬(Young-Ryul Kim),한창수(Chang-Soo Han) 제어로봇시스템학회 2010 제어·로봇·시스템학회 논문지 Vol.16 No.7

This paper presents an optimum tire force distribution method for 6WD/6WS(6-Wheel-Drive and 6-Wheel-Steering) electric vehicles. Using an independent steering and driving system, the performance of 6WD/6WS vehicles can be improved, as, for example, with respect to their maneuverability under low speed and their stability at high speed. Therefore, there should be a control strategy for finding the optimum tire forces that satisfy the driver’s command and minimize energy consumption. From the driver’s commands (steering angle and accelerator/brake pedal stroke), the desired yaw moment, the desired lateral force, and the desired longitudinal force were obtained. These three values were distributed to each wheel as the torque and the steering angle, based on the optimum tire force distribution method. The optimum tire force distribution method finds the longitudinal/lateral tire forces of each wheel that minimize the cost function, which is the sum of the normalized tire forces. Next, the longitudinal/lateral tire forces of each wheel are converted into the reference torque inputs and the steering wheel angle inputs. The proposed method was tested through a simulation, and its effectiveness was verified.

비홀로노믹 이동로봇의 자율주행을 위한 기하학적 경로 추종 및 장애물 회피 방법

김동형(Dong-Hyung Kim),김창준(Chang-Jun Kim),한창수(Chang-Soo Han) 제어로봇시스템학회 2010 제어·로봇·시스템학회 논문지 Vol.16 No.8

This paper presents a method that integrates the geometric path tracking and the obstacle avoidance for nonholonomic mobile robot. The mobile robot follows the path by moving through the turning radius given from the pure pursuit method which is the one of the geometric path tracking methods. And the obstacle generates the obstacle potential, from this potential, the virtual force is obtained. Therefore, the turning radius for avoiding the obstacle is calculated by proportional to the virtual force. By integrating the turning radius for avoiding the obstacle and the turning radius for following the path, the mobile robot follows the path and avoids the obstacle simultaneously. The effectiveness of the proposed method is verified through the real experiments for path tracking only, static obstacle avoidance, dynamic obstacle avoidance.

B-spline 곡선을 이용한 양팔 로봇의 궤적 계획

김동형(Dong-Hyung Kim),최윤성(Youn-Sung Choi),김현국(Hyeonguk Kim),전명재(Myoungjae Jun),백성훈(Sunghoon Back),이용석(Yongseok Lee),이지영(Ji Yeong Lee),한창수(Chang-Soo Han) 제어로봇시스템학회 2017 제어·로봇·시스템학회 논문지 Vol.23 No.3

This paper proposes an algorithm for the trajectory planning of a dual-arm robot. Since the original collision-free path obtained in a probabilistic manner is unnecessarily jerky, it may induce high acceleration in the robot’s actuators, so we regard this as a low-quality path. We propose a B-spline trajectory planning algorithm that transforms the path into a collision-free and smooth trajectory to improve its quality. We applied our proposed algorithm to a cell-phone assembly task using a dual-arm robot as a specific application example. The simulation results show that the smooth and collision-free trajectories can be generated for the task with our motion planning simulator.

H.264에서의 시방향(時方向) 에러은닉 기법

김동형,정제창,Kim, Dong-Hyung,Jeong, Je-Chang 한국통신학회 2005 韓國通信學會論文誌 Vol.30 No.2C

H.264/AVC 동영상 부호화 표준에서는 새로운 부호화 도구들이 추가되었으며 이를 통하여 보다 높은 압축 효율을 보인다. 추가된 부호화 도구들로 인하여 H.264내의 매크로블록은 이전의 부호화 표준들에 비하여 보다 많은 정보를 포함한다. H.264 부호화 표준에서는 하나의 매크로블록이 최대 16개까지의 서로 다른 움직임벡터를 가질 수 있으며 최대 4개의 서로 다른 참조프레임을 가질 수 있다. 또한 다양한 블록크기로의 움직임 추정하며 이는 매크로블록 모드로서 정의된다. 따라서 H.264내의 매크로블록은 기존보다 많은 움직임벡터를 가질 뿐만 아니라 기존에는 없던 참조프레임과 매크로블록 모드의 정보를 새로이 포함하고 있다. 본 논문은 H.264 부호화 표준의 특성을 고려하여 매크로블록이 포함하고 있는 다양한 크기의 움직임 벡터 및 참조 프레임 뿐 아니라 주변 매크로 블록들의 모드를 이용하여 서로 다른 블록 크기로 에러를 은닉함으로써 매크로블록이 포함하고 있는 정보를 최대한 활용하고 이를 통한 효율적인 에러은닉 알고리듬을 제안한다. 제안하는 알고리듬은 기존의 시방향 에러은닉 기법과 비교하여 적은 연산양만을 가지면서도 비트율에 관계없이 향상된 객관적 화질을 나타낸다. For the improvement of coding efficiency, the H.264 standard uses new coding tools. Using these coding tools, H.264 has achieved significant improvements from rate-distortion point of view. The adoption of these tools enables a macroblock in H.264 to have more information, sixteen motion vectors, four reference frames and a macroblock mode. In this paper, we present an efficient temporal error concealment algorithm by using not only motion vectors and reference frames but also macroblock mode of neighbor macroblocks. Our algorithm conceals the macroblock error with variable sizes, $16{\times}16,\;16{\times}8,\;8{\times}16,\;8{\times}8$ depending on the macroblock modes of neighbor macroblocks. Simulation results show that the proposed method increase the objective quality regardless of bit-rate and block error rate.