지문의 특징점을 추출하기 위한 개선된 세선화 알고리즘

이세연(Lee Seyeun),김용학(Kim Yonghak),이명호(Lee Myungho),이주신(Lee Jooshin) 한국정보기술학회 2006 한국정보기술학회논문지 Vol.4 No.5

In this paper, an enhanced thining algorithm which extracts feature parameters into the fingerprint image is proposed. The proposed algorithm is diveded between thining process and extracting feature parameters. In the thinning and extracting feature parameters into the fingerprint image, we proposed a algorithm which used serching-mask. The proposed algorithm is added existing algorithm as sub-iteration, the algorithm maintain the skeleton connection, cancel the noise branch and make a 1 pixel line into thicker than 2 pixels line. To verify the propriety of the proposed algorithm, we compared existing algorithm with proposed algorithm to examine connection and similarity.

X-ray 의료진단기를 위한 무선트리거발생기

이채철(Jaecheol Lee),고대식(Daesik Ko),이주신(Jooshin Lee),박종구(JongGoo Park) 한국정보기술학회 2007 Proceedings of KIIT Conference Vol.2007 No.-

본 연구에서는 x-ray 이미지 획득장치 장치와 X-ray 발생기 사이에 상호 동기를 무선방식으로 실현하여 기존의 장비사이의 호환성 문제와 설치상의 문제점을 극복하고, 촬영 시 실제 발생되는 x-ray 신호의 정확한 시점에서 이미지를 획득하여 보다 개선된 영상을 얻을 수 있도록 연구하여 보았다. 본 연구를 구현하기 위하여 X-ray 발생기로부터의 X-ray 신호레벨을 원격 감지하여 디지털 x-ray 이미지 획득장치에 필요한 Trigger pulse ACQ_START 신호와 ACQ_END 신호를 발생시키는 트리거 발생기를 설계 제작하고 그 성능을 연구분석 하였다. In this paper, we propose a method of trigger pulse generation to capture the image on time by making a synchronization between the x-ray generator and digital x-ray image acquisition system. we designed a wireless trigger pulse generation circuit to make a synchronization between x-ray generator and digital image acquisition system and analysis its performance. The X -ray radiated from x-ray generator is detected by this method using Geigertube. When it starts to detect a certain level of x-rayradiation or above from the air, this method starts to generate a ACQ_START signal to indicate the timing for image acquisition starting from digital image acquisition system. Hence, when it starts to detect under certain level of x-ray signal from the air, this method starts to generate a ACC_END signal to indicate the timing for image acquisition stop from digital image acquisition system. Image acquisition is activated only this time between ACQ_START and ACQ_END signal. By doing this wireless detecting of x-ray signal from remote, we can get more accurate timing for capturing the x-ray image and any type of x-ray generator can be connected to digital image acquisition system, regards of wired protocol. This makes easy installation.

트래픽 정보 취득을 위한 고속이동물체 속도 측정

이주신,Lee Jooshin 한국통신학회 2004 韓國通信學會論文誌 Vol.29 No.11C

본 논문은 트래픽 정보취득을 위하여 영상의 라인 샘플링을 이용한 고속이동물체 속도 측정 알고리즘을 제안하였다. 이동물체의 트래픽 정보 취득을 위한 속도 측정은 도로에 제 1 샘플라인과 제 2 샘플라인을 설정해 놓고, 물체가 샘플라인을 통과할 때 취득된 영상의 시변환 색조 데이터와 기준영상 색조 데이터 사이에서 차영상 기법으로 자동차를 검출하고, 자동차가 두 샘플라인 사이에 거리를 통과할 때 점유하는 프레임수로 속도를 측정하였다. 제 1 샘플라인과 제 2 샘플라인에서 각각 검출된 자동차의 색조로 동일성 판별을 하였다. 제안된 방법의 타당성을 검토하기 위하여 주행하는 자동차를 대상으로 동일성 판별 및 속도 측정을 한 결과, 동일성 판별은 두 개의 샘플링 라인을 통과하는 자동차의 색조 데이터로 판별됨을 보였고, 자동차의 속도 측정은 X-밴드 속도 측정 시스템과 비교한 결과 3% 이내임을 보였다. In this paper, velocity measurement of fast moving object for traffic information acquisition using line sampling of image is proposed. Velocity measurement for traffic information acquisition of moving object is that the first sample line and second sample line on the road is set, then car is detected by using difference image method between time-variance hue data of image when car is passing two sample lines and hue data of the reference image, and velocity of the car is measured by using frame number of video which is occupied by two sample lines. Identification of the car is performed by hue of the detected car between the first sample line and second sample line, respectively To examine the propriety of the proposed algorithm, identification and velocity measurement for driving car is evaluated. The evaluated results is that it is identified by hue data of car passing two sample lines, and the velocity measurement for driving car is less than 3% comparing with X-band speed gun.

배경 초기화 알고리즘에 의한 이동물체 검출

임강모(KangMo Lim),이주신(JooShin Lee) 한국정보기술학회 2005 한국정보기술학회논문지 Vol.3 No.2

In this paper, moving object detection based on background initialization algorithm is proposed. The proposed algorithm is that frames of sampled background image sequence are divided by block then the sum of graylevel value for each block pixel is calculated, respectively. The initialization of background image is performed by malting the maximum reference image and the minimum reference image which are recomposed by selecting only the maximum graylevel value and the minimum graylevel value in the sum of graylevel value of blocks located at the same position at each frame of the background image sequence. The current image frame is divided by block then the sum of graylevel value for each block pixel is calculated. Moving object detection is performed that if the calculated value is in the range of the graylevel value of two reference images, it is considered background, otherwise it is considered moving object. To examine the propriety of the proposed algorithm, the evaluated results for human in outdoor environments is that the error rate of the proposed method is less than the existing methods and the detection rate of the proposed method is better than the existing methods.

지능형 교통 시스템을 위한 이동물체 실시간 속도 측정

임강모(KangMo Lim),이주신(JooShin Lee) 한국정보기술학회 2005 한국정보기술학회논문지 Vol.3 No.3

In this paper. real-time velocity measurement of moving object tracking for intelligent transport system is proposed. A moving direction initialization of moving object is performed by obtaining moving angle and moving distance of moving objects which are detected at successive two frames of image sequence. The moving direction of the moving object at the current frame is obtained by applying the decimation-in-time search direction mask. A moving direction prediction of the moving object is performed by deciding the decimation-in-time search direction mask in the 8 search directions. The velocity measurement of moving object is perform어 by predicting the moving direction of the moving object as frames in the image sequence are decimated by applying the decimation-in-time search direction mask. To examine the propriety of the proposed algorithm, velocity measurement for driving car is evaluated. The evaluated results show that velocity of the car is measured at every 3 frames or 4 frames, and the processing time of the velocity measurement is 26.7㎳ on the average. The comparison of results of the proposed algorithm and results of X-band speed gun and the line sampling method is that the error is less than 1%.

이동물체 실시간 추적을 위한 시간솎음 탐색 알고리즘

임강모(KangMo Lim),이주신(JooShin Lee) 한국정보기술학회 2005 Proceedings of KIIT Conference Vol.2005 No.-

본 논문에서는 이동물체 실시간 추적을 위한 시간솎음 탐색 알고리즘을 제안하였다. 제안한 방법은 이동물체가 고속으로 움직일 경우에는 극히 짧은 시간 동안 이동 방향을 변경하기 어렵다는 가정 하에 이동물체의 이동 방향 초기화에 의해 이동 방향을 예측하고, 시간솎음 탐색 방향 마스크를 적용하여 프레임을 솎아냄으로써 이동물체 추적 처리 시간을 단축하는 방법으로써, 이동 방향 초기화는 영상 시퀀스의 연속된 프레임에서 검출된 이동물체의 이동 각도와 이동 거리를 구하여 수행하였고, 현재 프레임에서 이동물체의 이동 방향은 시간솎음 탐색 방향 마스크를 적용하여 구하였다. 이동물체의 이동 방향 예측은 이동물체의 이동 각도를 구하여 8 탐색 방향 중에서 시간솎음 탐색 방향 마스크를 결정하였고, 시간솎음 탐색 방향 마스크를 적용하여 프레임을 시간 솎음하고 이동물체의 이동 방향을 예측하여 이동물체를 추적하였다. 제안한 방법의 타당성을 입증하기 위하여 주행 중인 자동차를 대상으로 추적 실험을 수행한 결과, 제안한 방법은 전역탐색기법에 비해서 이동물체 변위 탐색 횟수가 평균 92% 감소하였고, 이동물체 추적 처리 시간은 평균 26.7㎳가 걸림으로서 실시간 처리가 가능함을 보였다. In this paper. decimation-in-time search algorithm for real-time tracking of moving object is proposed. A moving direction initialization of moving object is performed by obtaining moving angle and moving distance of moving objects which are detected at successive two frames of image sequence. The moving direction of the moving object at the current frame is obtained by applying the decimation-in-time search direction mask. A moving direction prediction of the moving object is performed by deciding the decimation-in-time search direction mask in the 8 search directions. The moving object tracking is performed by predicting the moving direction of the moving object as frames in the image sequence are decimated by applying the decimation-in-time search direction mask. To examine the propriety of the proposed algorithm, the moving object tracking is simulated for driving car. The results of the moving object tracking show that the number of displacement search times is reduced up to 92% on the average in the proposed algorithm, and the processing time of the tracking is 37.4ms on the average, and hence it is possible to perform real-time tracking of moving object.