• Title, Summary, Keyword: 영상처리

Search Result 9,049, Processing Time 0.084 seconds

3D Facial Animation with Head Motion Estimation and Facial Expression Cloning (얼굴 모션 추정과 표정 복제에 의한 3차원 얼굴 애니메이션)

  • Kwon, Oh-Ryun;Chun, Jun-Chul
    • The KIPS Transactions:PartB
    • /
    • v.14B no.4
    • /
    • pp.311-320
    • /
    • 2007
  • This paper presents vision-based 3D facial expression animation technique and system which provide the robust 3D head pose estimation and real-time facial expression control. Many researches of 3D face animation have been done for the facial expression control itself rather than focusing on 3D head motion tracking. However, the head motion tracking is one of critical issues to be solved for developing realistic facial animation. In this research, we developed an integrated animation system that includes 3D head motion tracking and facial expression control at the same time. The proposed system consists of three major phases: face detection, 3D head motion tracking, and facial expression control. For face detection, with the non-parametric HT skin color model and template matching, we can detect the facial region efficiently from video frame. For 3D head motion tracking, we exploit the cylindrical head model that is projected to the initial head motion template. Given an initial reference template of the face image and the corresponding head motion, the cylindrical head model is created and the foil head motion is traced based on the optical flow method. For the facial expression cloning we utilize the feature-based method, The major facial feature points are detected by the geometry of information of the face with template matching and traced by optical flow. Since the locations of varying feature points are composed of head motion and facial expression information, the animation parameters which describe the variation of the facial features are acquired from geometrically transformed frontal head pose image. Finally, the facial expression cloning is done by two fitting process. The control points of the 3D model are varied applying the animation parameters to the face model, and the non-feature points around the control points are changed by use of Radial Basis Function(RBF). From the experiment, we can prove that the developed vision-based animation system can create realistic facial animation with robust head pose estimation and facial variation from input video image.

Influence of Gravel Content and Nitrogen Application on Nitrogen Leaching by the Leachate and Chinese Cabbage Growth in Highland (자갈함량과 질소시비량이 고랭지 배추재배시 침투수에 의한 질소용탈 및 생육에 미치는 영향)

  • Park, Chol-Soo;Lee, Gye-Jun;Jung, Yeong-Sang;Joo, Jin-Ho;Hwang, Seon-Woong
    • Korean Journal of Environmental Agriculture
    • /
    • v.24 no.1
    • /
    • pp.1-5
    • /
    • 2005
  • Continuous monocropping of Chinese cabbage in Gangwon highland increased gravel and sand contents due to surface soil erosion. Nutrient leaching and Chinese cabbage growth were investigated with different treatments of gravel contents and nitrogen application levels by using $0.5m^2$ Wagner pots. Gravel contents were 0, 10, 30, 50, 70, and 90%(w/w), nitrogen application levels were 60, 120, and 240 kg/ha, and manure compost application rate was 15 ton per hectare, respectively. Wagner pots were filled with loamy sand soil mixed with 5 cm-sized gravels. Fresh weight of Chinese cabbage was decreased as gravel contents in soil increased, and particularly severely decreased at 240 kg-N/ha. Yields of Chinese cabbage were remarkably decreased at the rate of 60 kg-N/ha with 30% gravel content and 120 kg-N/ha with 50% gravel content. Most of Chinese cabbages were severely wilted by heavy N application at the rate of 240 kg-N/ha in the middle of growth stages regardless of gravel contents, while about 50% of Chinese cabbage showed wilting symptom in the treatment of more than 50% of gravel contents and 120 kg-N/ha. N content in leachate increased as gravel content and N application increased. The relationship between gravel content and N contents showed linear regression: N in leachate = 0.014(gravel content) -0.039 (r = 0.961). Particularly, $NH_4-N$ contents in leachates with more than 30% gravel content and 240 kg-N/ha ranged from $139{\sim}339mg/L$. Chinese cabbage growth in loamy sand soil containing 30%, and 50% gravel contents could be adversely affected by N application at the rate of 240, and 120 kg-N/ha, respectively.

Object Tracking Based on Exactly Reweighted Online Total-Error-Rate Minimization (정확히 재가중되는 온라인 전체 에러율 최소화 기반의 객체 추적)

  • JANG, Se-In;PARK, Choong-Shik
    • Journal of Intelligence and Information Systems
    • /
    • v.25 no.4
    • /
    • pp.53-65
    • /
    • 2019
  • Object tracking is one of important steps to achieve video-based surveillance systems. Object tracking is considered as an essential task similar to object detection and recognition. In order to perform object tracking, various machine learning methods (e.g., least-squares, perceptron and support vector machine) can be applied for different designs of tracking systems. In general, generative methods (e.g., principal component analysis) were utilized due to its simplicity and effectiveness. However, the generative methods were only focused on modeling the target object. Due to this limitation, discriminative methods (e.g., binary classification) were adopted to distinguish the target object and the background. Among the machine learning methods for binary classification, total error rate minimization can be used as one of successful machine learning methods for binary classification. The total error rate minimization can achieve a global minimum due to a quadratic approximation to a step function while other methods (e.g., support vector machine) seek local minima using nonlinear functions (e.g., hinge loss function). Due to this quadratic approximation, the total error rate minimization could obtain appropriate properties in solving optimization problems for binary classification. However, this total error rate minimization was based on a batch mode setting. The batch mode setting can be limited to several applications under offline learning. Due to limited computing resources, offline learning could not handle large scale data sets. Compared to offline learning, online learning can update its solution without storing all training samples in learning process. Due to increment of large scale data sets, online learning becomes one of essential properties for various applications. Since object tracking needs to handle data samples in real time, online learning based total error rate minimization methods are necessary to efficiently address object tracking problems. Due to the need of the online learning, an online learning based total error rate minimization method was developed. However, an approximately reweighted technique was developed. Although the approximation technique is utilized, this online version of the total error rate minimization could achieve good performances in biometric applications. However, this method is assumed that the total error rate minimization can be asymptotically achieved when only the number of training samples is infinite. Although there is the assumption to achieve the total error rate minimization, the approximation issue can continuously accumulate learning errors according to increment of training samples. Due to this reason, the approximated online learning solution can then lead a wrong solution. The wrong solution can make significant errors when it is applied to surveillance systems. In this paper, we propose an exactly reweighted technique to recursively update the solution of the total error rate minimization in online learning manner. Compared to the approximately reweighted online total error rate minimization, an exactly reweighted online total error rate minimization is achieved. The proposed exact online learning method based on the total error rate minimization is then applied to object tracking problems. In our object tracking system, particle filtering is adopted. In particle filtering, our observation model is consisted of both generative and discriminative methods to leverage the advantages between generative and discriminative properties. In our experiments, our proposed object tracking system achieves promising performances on 8 public video sequences over competing object tracking systems. The paired t-test is also reported to evaluate its quality of the results. Our proposed online learning method can be extended under the deep learning architecture which can cover the shallow and deep networks. Moreover, online learning methods, that need the exact reweighting process, can use our proposed reweighting technique. In addition to object tracking, the proposed online learning method can be easily applied to object detection and recognition. Therefore, our proposed methods can contribute to online learning community and object tracking, detection and recognition communities.

Risk Factor Analysis for Operative Death and Brain Injury after Surgery of Stanford Type A Aortic Dissection (스탠포드 A형 대동맥 박리증 수술 후 수술 사망과 뇌손상의 위험인자 분석)

  • Kim Jae-Hyun;Oh Sam-Sae;Lee Chang-Ha;Baek Man-Jong;Hwang Seong-Wook;Lee Cheul;Lim Hong-Gook;Na Chan-Young
    • Journal of Chest Surgery
    • /
    • v.39 no.4
    • /
    • pp.289-297
    • /
    • 2006
  • Background: Surgery for Stanford type A aortic dissection shows a high operative mortality rate and frequent postoperative brain injury. This study was designed to find out the risk factors leading to operative mortality and brain injury after surgical repair in patients with type A aortic dissection. Material and Method: One hundred and eleven patients with type A aortic dissection who underwent surgical repair between February, 1995 and January 2005 were reviewed retrospectively. There were 99 acute dissections and 12 chronic dissections. Univariate and multivariate analysis were performed to identify risk factors of operative mortality and brain injury. Resuit: Hospital mortality occurred in 6 patients (5.4%). Permanent neurologic deficit occurred in 8 patients (7.2%) and transient neurologic deficit in 4 (3.6%). Overall 1, 5, 7 year survival rate was 94.4, 86.3, and 81.5%, respectively. Univariate analysis revealed 4 risk factors to be statistically significant as predictors of mortality: previous chronic type III dissection, emergency operation, intimal tear in aortic arch, and deep hypothemic circulatory arrest (DHCA) for more than 45 minutes. Multivariate analysis revealed previous chronic type III aortic dissection (odds ratio (OR) 52.2), and DHCA for more than 45 minutes (OR 12.0) as risk factors of operative mortality. Pathological obesity (OR 12.9) and total arch replacement (OR 8.5) were statistically significant risk factors of brain injury in multivariate analysis. Conclusion: The result of surgical repair for Stanford type A aortic dissection was good when we took into account the mortality rate, the incidence of neurologic injury, and the long-term survival rate. Surgery of type A aortic dissection in patients with a history of chronic type III dissection may increase the risk of operative mortality. Special care should be taken and efforts to reduce the hypothermic circulatory arrest time should alway: be kept in mind. Surgeons who are planning to operate on patients with pathological obesity, or total arch replacement should be seriously consider for there is a higher risk of brain injury.

Comparative study on quality of scanned images from varying materials and surface conditions of standardized model for dental scanner evaluation (치과용 스캐너 평가를 위한 국제표준모델의 재료 및 표면 상태에 따른 스캔 영상 결과물 비교 연구)

  • Park, Ju-Hee;Seol, Jeong-Hwan;Lee, Jun Jae;Lee, Seung-Pyo;Lim, Young-Jun
    • Journal of Dental Rehabilitation and Applied Science
    • /
    • v.34 no.2
    • /
    • pp.104-115
    • /
    • 2018
  • Purpose: The purpose of this study is to evaluate the image acquisition ability of intraoral scanners by analyzing the comprehensiveness of scanned images from standardized model, and to identify problems of the model. Materials and Methods: Cast models and 3D-printed models were prepared according to international standards set by ISO12836 and ANSI/ADA no. 132, which were then scanned by model scanner and two different intraoral scanners (TRIOS3 and CS3500). The image acquisition performance of the scanners was classified into three grades, and the study was repeated with varying surface conditions of the models. Results: Model scanner produced the most accurate images in all models. Meanwhile, CS3500 showed good image reproducibility for angled structures and TRIOS3 showed good image reproducibility for rounded structures. As for model ingredients, improved plaster model best reproduced scan images regardless of the type of scanner used. When limited to 3D-printed model, powdered surface condition resulted in higher image quality. Conclusion: When scanning structures beyond FOV (field of view) in standardized models (following ISO12836 and ANSI/ADA 132), lack of reference points to help distinguish different faces confuses the scanning and matching process, resulting in inaccurate display of images. These results imply the need to develop a new standard model not confined to simple pattern repetition and symmetric structure.

Dynamic Traffic Assignment Using Genetic Algorithm (유전자 알고리즘을 이용한 동적통행배정에 관한 연구)

  • Park, Kyung-Chul;Park, Chang-Ho;Chon, Kyung-Soo;Rhee, Sung-Mo
    • Journal of Korean Society for Geospatial Information Science
    • /
    • v.8 no.1
    • /
    • pp.51-63
    • /
    • 2000
  • Dynamic traffic assignment(DTA) has been a topic of substantial research during the past decade. While DTA is gradually maturing, many aspects of DTA still need improvement, especially regarding its formulation and solution algerian Recently, with its promise for In(Intelligent Transportation System) and GIS(Geographic Information System) applications, DTA have received increasing attention. This potential also implies higher requirement for DTA modeling, especially regarding its solution efficiency for real-time implementation. But DTA have many mathematical difficulties in searching process due to the complexity of spatial and temporal variables. Although many solution algorithms have been studied, conventional methods cannot iud the solution in case that objective function or constraints is not convex. In this paper, the genetic algorithm to find the solution of DTA is applied and the Merchant-Nemhauser model is used as DTA model because it has a nonconvex constraint set. To handle the nonconvex constraint set the GENOCOP III system which is a kind of the genetic algorithm is used in this study. Results for the sample network have been compared with the results of conventional method.

  • PDF

Radiation Dose Accuracy 81 the Isocenter : Standard Stereotactic Radiosurgery Technique Developed at Seoul National University Hospital (서울대학교병원형 방사선수술 표준기법의 중심점 선량의 오차)

  • Shin Seong Soo;Kim Il Han;Ha Sung Whan;Park Charn Il;Kang Wee-Saing;Hur Sun Nyung
    • Radiation Oncology Journal
    • /
    • v.20 no.4
    • /
    • pp.391-395
    • /
    • 2002
  • Purpose : To confirm the accuracy of the radiation dose at the isocenter by the standard linear accelerator-based stereotactic radiosurgery technique which was developed at Seoul National University Hospital. Materials and Methods : Radiation dosimetry was undertaken during standard 5-arc radiosurgery using 6 MV X-ray beam from CL2100C linac. The treatment head was attached with circular tertiary collimators of 10 and 20 mm diameter. We measured the absorbed dose at the isocenter of a multi-purpose phantom using two kinds of detector : a 0.125 co ionization chamber and a silicon diode detector. Results : The dose differences at each arc plane between the planned dose and the measured dose at the isocenter raged from $-0.73\%\;to\;-2.69\%$ with the 0.125 cc ion chamber, and from $-1.29\%\;to\;-2.91\%$ with the diode detector during radiosurgery with the tertiary collimator of 20 mm diameter. Those with the 10-mm tertiary collimator ranged from $-2.39\%\;to\;-4.25\%$ with the diode. Conclusion : The dose accuracy at the isocenter was ${\pm}3\%$. Therefore, further efforts such ws modification in processing of the archived image through DICOM3.0 format are required to lessen the dose difference.

Design of Translator for generating Secure Java Bytecode from Thread code of Multithreaded Models (다중스레드 모델의 스레드 코드를 안전한 자바 바이트코드로 변환하기 위한 번역기 설계)

  • 김기태;유원희
    • Proceedings of the Korea Society for Industrial Systems Conference
    • /
    • /
    • pp.148-155
    • /
    • 2002
  • Multithreaded models improve the efficiency of parallel systems by combining inner parallelism, asynchronous data availability and the locality of von Neumann model. This model executes thread code which is generated by compiler and of which quality is given by the method of generation. But multithreaded models have the demerit that execution model is restricted to a specific platform. On the contrary, Java has the platform independency, so if we can translate from threads code to Java bytecode, we can use the advantages of multithreaded models in many platforms. Java executes Java bytecode which is intermediate language format for Java virtual machine. Java bytecode plays a role of an intermediate language in translator and Java virtual machine work as back-end in translator. But, Java bytecode which is translated from multithreaded models have the demerit that it is not secure. This paper, multhithread code whose feature of platform independent can execute in java virtual machine. We design and implement translator which translate from thread code of multithreaded code to Java bytecode and which check secure problems from Java bytecode.

  • PDF

Using Spatial Data and Crop Growth Modeling to Predict Performance of South Korean Rice Varieties Grown in Western Coastal Plains in North Korea (공간정보와 생육모의에 의한 남한 벼 품종의 북한 서부지대 적응성 예측)

  • 김영호;김희동;한상욱;최재연;구자민;정유란;김재영;윤진일
    • Korean Journal of Agricultural and Forest Meteorology
    • /
    • v.4 no.4
    • /
    • pp.224-236
    • /
    • 2002
  • A long-term growth simulation was performed at 496 land units in the western coastal plains (WCP) of North Korea to test the potential adaptability of each land unit for growing South Korean rice cultivars. The land units for rice cultivation (CZU), each of them represented by a geographically referenced 5 by 5 km grid tell, were identified by analyzing satellite remote sensing data. Surfaces of monthly climatic normals for daily maximum and minimum temperature, precipitation number of rain days and solar radiation were generated at a 1 by 1 km interval by spatial statistical methods using observed data at 51 synoptic weather stations in North and South Korea during 1981-2000. Grid cells felling within a same CZU and, at the same time, corresponding to the satellite data- identified rice growing pixels were extracted and aggregated to make a spatially explicit climatic normals relevant to the rice growing area of the CZU. Daily weather dataset for 30 years was randomly generated from the monthly climatic normals of each CZU. Growth and development parameters of CERES-rice model suitable for 11 major South Korean cultivars were derived from long-term field observations. Eight treatments comprised of 2 transplanting dates $\times$ 2 cropping systems $\times$ 2 irrigation methods were assigned to each cultivar. Each treatment was simulated with the randomly generated 30 years' daily weather data (from planting to physiological maturity) for 496 land units in WCP to simulate the growth and yield responses to the interannual climate variation. The same model was run with the input data from the 3 major crop experiment stations in South Korea to obtain a 30 year normal performance of each cultivar, which was used as a "reference" for comparison. Results were analyzed with respect to spatial and temporal variation in yield and maturity, and used to evaluate the suitability of each land unit for growing a specific South Korean cultivar. The results may be utilized as decision aids for agrotechnology transfer to North Korea, for example, germplasm evaluation, resource allocation and crop calendar preparation.

Influence of Soil Texture and Bulk Density on Root Growth Characteristics and Nutrient Influx Rate of Soybean Plant (토성(土性)과 용적밀도(容積密度)가 대두(大豆)의 뿌리 생장특성(生長特性)과 양분흡수기능(養分吸收機能)에 미치는 영향(影響))

  • Jung, Yeong-Sang;Lim, Hyung-Sik
    • Korean Journal of Soil Science and Fertilizer
    • /
    • v.22 no.3
    • /
    • pp.221-227
    • /
    • 1989
  • This study was conducted to understand the influence of soil compaction on root growth and nutrient uptake characteristics of the soybean roots grown in two soils with different texture. Tap root elongation was measured on young seedling grown in cores compacted to different bulk densities of 1.2, 1.4 and $1.6/cm^3$ with different soil water retention in laboratory. The soil used were Samgag sandy loam and Baegsan loam soils. The wet and dry weight, total length, average radius and total surface area of roots were measured on soybean plants grown in 1/5000 a Wagner pots compacted to different bulk density of 1.2 and $1.4g/cm^3$. The nutrient uptake of soybean shoot was measured and evaluated with the unit surface area of roots at the 7th, 17th and 27th days after germination. The results were as follows: 1. The tap root elongation rate was faster in the loam soil with low bulk density than in the sandy loam soil with high bulk density. The elongation rates were remarkedly decreased when soil water was lower than the retention of 4 bars in loam soil and that of 1 bars in sandy loam soil. 2. Tap root elongation rate sharply decreased as increased soil strength higher than $2kgf/cm^2$ measured by ELE penetrometer showing curvillinear regression. However, it was low regardless of soil strength when soil water retention was 10 bars in sandy loam soil. 3. From the pot experiment, the total length of roots were longer in loam soil than in sandy loam soil and was longer in the soils with lower bulk density. The average radius of fine roots grown in sandy loam soil was larger than that grown in loam soil. The total surface area of roots was greater in the loam soil with low bulk density than in the sandy loam soil with high bulk density as the total length of roots. 4. The amounts of nutrient uptake by soybean shoots were greater in loam soil primarily due to more production of dry matter than in sandy loam soil. The nitrogen influx rates through the unit surface area were 597 to $753nmoles/day-cm^2$ in loam soil and 222 to $365nmoles/day\;cm^2$ in sandy loam soilshowing higher value in higher bulk density. The potasium influx rates were 99 to $175nmoles/day-cm^2$, and those of phosphate were 26 to $46nmoles/day\;cm^2$. Those of Ca and Mg were 175 to 246 and 163 to $205nmoles/day\;cm^2$. The difference in nutrient influx rates between bulk densities of these elements were lower than that of nitrogen.

  • PDF