• Title/Summary/Keyword: MEAN SHIFT

Search Result 645, Processing Time 0.034 seconds

A design of MPEG-4 video object segmentation using color/motion information (칼라/움직임 정보를 이용한 MPEG-4 비디오 객체 분할 설계)

  • 김준기;이호석
    • Proceedings of the Korean Information Science Society Conference
    • /
    • 2000.10b
    • /
    • pp.206-208
    • /
    • 2000
  • 본 논문은 칼라 정보와 움직임 정보를 이용한 객체 분할 기법의 설계에 대하여 소개한다. 객체 분할 알고리즘은 L*u*v 공간의 칼라 특성과 움직임 특성을 결합하여 설계하였다. 즉 공간 분할은 mean shift 칼라 클러스터링 알고리즘(color clustering algorithm)을 사용하여 중심 칼라 영역에 따라 동일한 칼라 지역으로 통합한다. 시간 분할은 움직임 검출을 위하여 affine six parameter 움직임 모델과 optical flow equation를 이용하여 움직임이 발생한 부분을 검출한다. 다음에 공간 분할과 시간 분할에 따라 결과를 통합하고 MAD(mean absolute difference)를 사용하여 객체를 추출하는 알고리즘을 설계하였다.

  • PDF

The Effect of Shift-Working Nurses' Need Satisfaction, Calling and Vocation, and Job Stress on Retention Intention (교대근무간호사의 욕구만족, 소명의식, 직무스트레스가 재직의도에 미치는 영향)

  • Eu-Nam Lee;Hwee Wee
    • Korea Journal of Hospital Management
    • /
    • v.29 no.2
    • /
    • pp.1-17
    • /
    • 2024
  • Purpose: This descriptive study aimed to identify shift nurses' retention intention, need satisfaction, calling and vocation, and job stress, as well as the effects of these factors on retention intention. Methods: The participants of this study were 193 nurses working shifts with more than three months' clinical experience at general hospitals. Data were collected using a structured questionnaire from February 1 to 28, 2023. Nurses' Retention Index and Nurse Needs Satisfaction Scale, Korean version of the Calling and Vocational Questionnaire, Korean Occupational Stress Scale were used. Data were analyzed using SPSS/WIN 23.0 program. Independent t-test, one-way ANOVA, Pearson's correlation analysis, and hierarchical multiple regression analysis were conducted for statistical analysis. Results: Regarding nurses' shift work, the mean scores for retention intention, needs satisfaction, calling and vocation, and job stress were 32.12±9.04, 94.87±14.74, 26.95±6.84, and 66.12±10.34. Moreover, the nurses' retention intention differed significantly by age, education level, religion, marital status, clinical experience, satisfaction with wage, and job satisfaction. Furthermore, retention intention was positively correlated with needs satisfaction and calling and vocation and negatively correlated with job stress. Needs satisfaction, and calling and vocation as well as education level of bachelor's degree, job satisfaction had positive effects on shift work nurses' retention intention. Job dissatisfaction had relatively negative effects on retention intention compared to having neutral job satisfaction. Conclusion: It is necessary to develop programs that promote needs satisfaction, calling and vocation, and job satisfaction to improve nurses' retention intention in relation to shift work; moreover, institutional support for acquiring a bachelor's degree or higher is needed.

  • PDF

Asymptotic Properties of Outlier Tests in Nonlinear Regression

  • Kahng, Myung-Wook
    • Journal of the Korean Data and Information Science Society
    • /
    • v.17 no.1
    • /
    • pp.205-211
    • /
    • 2006
  • For a linear regression model, the necessary and sufficient condition for the asymptotic consistency of the outlier test statistic is known. An analogous condition for the nonlinear regression model is considered in this paper.

  • PDF

Spontaneous Resolution of Chronic Subdural Hematoma : Close Observation as a Treatment Strategy

  • Kim, Hyung Chan;Ko, Jung Ho;Yoo, Dong Soo;Lee, Sang-Koo
    • Journal of Korean Neurosurgical Society
    • /
    • v.59 no.6
    • /
    • pp.628-636
    • /
    • 2016
  • Objective : Chronic subdural hematoma (cSDH) is common condition in neurosurgical field. It is difficult to select the treatment modality between the surgical method and the conservative method when patients have no or mild symptoms. The purpose of this study is to provide a suggestion that the patients could be cured with conservative treatment modality. Methods : We enrolled 16 patients who had received conservative treatment for cSDH without special medications which could affect hematoma resolution such as mannitol, steroids, tranexamic acid and angiotensin converting enzyme inhibitors. The patients were classified according to the Markwalder's Grading Scale. Results : Among these 16 patients, 13 (81.3%) patients showed spontaneously resolved cSDH and 3 (18.7%) patients received surgery due to symptom aggravation and growing hematoma. They were categorized into two groups based on whether they were cured with conservative treatment or not. The first group was the spontaneous resolution group. The second group was the progression-surgery group. The mean hematoma volume in the spontaneous resolution group was 43.1 mL. The mean degree of midline shift in the spontaneous resolution group was 5.3 mm. The mean hematoma volume in the progression-surgery group was 62.0 mL. The mean degree of midline shift in the second group was 6 mm. Conclusion : We suggest that the treatment modality should be determined according to the patient's symptoms and clinical condition and close observation could be performed in patients who do not have any symptoms or in patients who have mild to moderate headache without neurological deterioration.

A Method for Tree Image Segmentation Combined Adaptive Mean Shifting with Image Abstraction

  • Yang, Ting-ting;Zhou, Su-yin;Xu, Ai-jun;Yin, Jian-xin
    • Journal of Information Processing Systems
    • /
    • v.16 no.6
    • /
    • pp.1424-1436
    • /
    • 2020
  • Although huge progress has been made in current image segmentation work, there are still no efficient segmentation strategies for tree image which is taken from natural environment and contains complex background. To improve those problems, we propose a method for tree image segmentation combining adaptive mean shifting with image abstraction. Our approach perform better than others because it focuses mainly on the background of image and characteristics of the tree itself. First, we abstract the original tree image using bilateral filtering and image pyramid from multiple perspectives, which can reduce the influence of the background and tree canopy gaps on clustering. Spatial location and gray scale features are obtained by step detection and the insertion rule method, respectively. Bandwidths calculated by spatial location and gray scale features are then used to determine the size of the Gaussian kernel function and in the mean shift clustering. Furthermore, the flood fill method is employed to fill the results of clustering and highlight the region of interest. To prove the effectiveness of tree image abstractions on image clustering, we compared different abstraction levels and achieved the optimal clustering results. For our algorithm, the average segmentation accuracy (SA), over-segmentation rate (OR), and under-segmentation rate (UR) of the crown are 91.21%, 3.54%, and 9.85%, respectively. The average values of the trunk are 92.78%, 8.16%, and 7.93%, respectively. Comparing the results of our method experimentally with other popular tree image segmentation methods, our segmentation method get rid of human interaction and shows higher SA. Meanwhile, this work shows a promising application prospect on visual reconstruction and factors measurement of tree.

Reconstruction of Anterior Cruciate Ligament with Fresh-frozen Achilles Allograft (신선 동결 동종 아킬레스건을 이용한 전방 십자인대 재건술)

  • Choi, Nam-Yong;Han, Suk-Ku;Nah, Ki-Ho;Seo, Sang-ll;Lee, Do-Sung;Song, Hyun-Seok
    • Journal of the Korean Arthroscopy Society
    • /
    • v.8 no.1
    • /
    • pp.26-30
    • /
    • 2004
  • Purpose: We analyzed the clinical results of ACL(anterior cruciate ligament) reconstruction of chronic anterior cruciate ligament injuries with a fresh frozen Achilles allograft, retrospectively. Methods: Thirty-eight patients(average age 32.4 years) who had an chronic injury of ACL reconstructed with an Achilles allograft between January 1999 and February 2001 were included. The mean follow-up was 19.2(12-36) months. The clinical evaluation was done by range of motion, Lachman and pivot-shift test, KT-1000 arthrometer, Lysholm knee score, and the modified Feagin score. Results: All patients recovered full range of motion of affected knees. With use of Lachman and pivot-shift test data, thirty seven(97.4%) were functional. When comparing KT-1000, we found mean maximum side-to-side differences less than 2.1 mm. According to Lysholm knee score, mean postoperative score was 90.4. Thirty four(89.5%) had good to excellent results in modified Feagin score. Conclusion: Results of ACL reconstruction using fresh frozen Achilles allografts were comparable to autografts. ACL reconstruction with Achilles allograft could be a reasonable altemative to autografts.

  • PDF

A Study on Glass Tile Generation for Stained Glass Rendering (스테인드 글라스 렌더링을 위한 유리 타일 생성에 관한 연구)

  • Nah, Hyeon-Cheol;Gi, Yong-Jea;Yoon, Kyung-Hyun
    • 한국HCI학회:학술대회논문집
    • /
    • 2006.02a
    • /
    • pp.1246-1251
    • /
    • 2006
  • 본 연구에서는 영역 분할 알고리즘과 3차 스플라인 보간법을 이용하여 스테인드 글라스 렌더링을 위한 개선된 유리 타일 생성 알고리즘을 제안하였다. 먼저 유리 타일의 초기 형태를 추출하기 위하여 입력 영상에 Mean shift 분할 알고리즘을 적용하였다. Mean shift 분할 알고리즘은 영상의 각 픽셀(pixel)에서의 지역 밀도 최대 점(local density maximum)을 찾아 클러스터링(clustering)하는 알고리즘으로 영상을 효과적으로 분할할 수 있다. 그리고 분할된 영역에서 영역을 사용자 입력으로 병합하고, 영역에서 부적절한 형태를 없애기 위해 본 연구에서는 형태론적 연산(morphological operation)을 이용하였다. 추출된 영역으로부터 유리 타일의 형태로 만들기 위하여 추출된 각각의 영역에 3차 스플라인 보간법(cubic spline interpolation)을 적용하여 경계가 완화된 영역과 납틀(leading)의 형태를 얻는다. 그 다음 영역을 스플라인 곡선(spline curve)을 이용하여 재분할하고, 각 영역에 변환(transformation)된 색상을 적용하여 최종적인 유리 타일을 만들어낸다. 본 연구에서는 3차 스플라인 보간법을 이용하여 실제 스테인드 글라스에서 생길 수 있는 부드러운 경계를 갖는 유리 타일의 형태를 만들어 이를 스테인드 글라스 렌더링에 이용하였다. 이 방법은 기존의 영역 분할 알고리즘에 형태론적 연산만을 적용하여 유리 타일의 형태를 생성하는 것보다 효과적으로 유리 타일의 형태를 생성할 수 있다. 또한, 생성된 영역에 재분할 과정을 거쳐서 작은 유리 타일이 모여서 이루는 조형적인 형태를 이룰 수 있도록 하였다.

  • PDF

A Study on the Effects of Immediate Side Shift to the Pantographic Reproducibility Index (Immediate Side Shift가 Pantographic Reproducibility Index에 끼치는 영향에 관한 연구)

  • Nam, Cheon Woo;Han, Kyung Soo
    • Journal of Oral Medicine and Pain
    • /
    • v.12 no.1
    • /
    • pp.75-83
    • /
    • 1987
  • This study was designed to investigate the effects of TMJ incoordination to condylar movements, especially, the ISS. The sounds are one of the symptoms in TMJ incoordinated disorder, and it may cause the changes of mandibular movement trajectory. 19 students with only TMJ sounds and 16 students with no TMJ problems participated in this study. The subject performed Rt. lateral, Lt. lateral and protrusive movements, and repeated 3 times on each movement. Pantronic was used to record the measures of condylar movement paths. The obtained results were as follows : 1. The mean values of RISS and LISS in control group were 0.29mm, 0.36mm respectively, and those in experimental group were 0.49mm, 0.41mm repectively. The mean values of RISS was higher in experimental group than that of RISS in control group. 2. Correlation coefficients between PRI and RISS, LISS were slightly higher in experimental group than those in control group, therefore, PRI was more likely to be affected by ISS in experimental group. 3. In control group PRI was correlated to RISS, LORB, RPRO and LPRO, but in experimental group PRI was not correlated to those items. From the study, the author knew that the condylar movements was stable in control group.

  • PDF

A Study on Rule-Based Vehicle Tracking in Video Images (비디오 영상에서 규칙기반 차량추적에 관한 연구)

  • Park Eun-Jong;Lee Joon-Whan
    • The Journal of The Korea Institute of Intelligent Transport Systems
    • /
    • v.4 no.2 s.7
    • /
    • pp.1-11
    • /
    • 2005
  • Automatic tracking of vehicles is important to accurately estimate the vehicle speeds in video-based traffic measurement systems and to analyze traffic flows for road construction. This paper proposes a carefully designed rule-based tracking scheme that considers the possible cases that can be appeared in the video-based vehicle racking. The proposed scheme is fast and outperforms the Mean-Shift scheme in terms of accuracy. The accuracy and the speed of the scheme would be increased by combining it with color-based searching and Kalman filters.

  • PDF