Abstract:
PURPOSE: A system for estimating a three-dimensional direction and a three-dimensional location and a method thereof are provided to reduce manufacturing costs by implementing a transmission unit and a reception unit based on ultraviolet rays. CONSTITUTION: A three-dimensional location or direction estimation system includes an estimator (351), a distance calculator (352), and a priority determining unit (353). An estimation unit estimates the direction of a remote device and the three-dimensional location of a remote device by using signal attenuation characteristics which are changed according to the direction and distance. The distance calculation unit calculates the distance between a transmission unit and a reception unit by using the three-dimensional location of the remote device. The priority determining unit determines the priority of the intensity information according to noise degrees by confirming the noise degree of the intensity information according to the direction of the remote device and the noise sensitivity of the intensity information according to the distance between the transmission unit and the reception unit. [Reference numerals] (310) Remote device; (320) Transmitter; (321) First transmission unit; (322) N^th transmission unit; (330) Receiver; (331) First reception unit; (332) Second reception unit; (334) N^th reception unit; (350) Corrector; (351) Estimator; (352) Distance calculator; (353) Priority determining unit; (AA) Intensity information
Abstract:
PURPOSE: A virtual world processing device and a method thereof are provided to apply a face expression of a user in the real world to a face expression of an avatar in a virtual world even if the expressions are different from each other. CONSTITUTION: An intelligent camera senses a face expression basis of a user. A receiving unit(210) receives the sensed information from the intelligent camera. Initialization information indicates a parameter initializing the face expression basis. A 2D/3D model designates a encoding method of a face object of an avatar. A processing unit(220) generates face expression data of the user. [Reference numerals] (210) Receiving unit; (220) Processing unit; (230) Output unit
Abstract:
PURPOSE: An electrode for a body and a device for measuring a body signal therewith are provided to reduce motion artifact by arranging a non-conductive adhesive sheet to be separated from the electrode. CONSTITUTION: An electrode(100) for a body comprises a first sheet member(10), a second sheet member(20), one or more metal contact points(50), and one or more electrodes(30). One or more holes(40) are formed on the first sheet member. The metal contact points are formed on the first sheet member. The electrodes are arranged on the second sheet member. First adhesive layers(15) are arranged on the top and bottom surfaces of the first sheet member.
Abstract:
인물 DB 등록사진 수 축소 방법 및 장치, 포토 앨범 시스템의 반자동등록 방법 및 장치와, 이를 이용한 포토앨범 시스템이 개시된다. 그 포토앨범 시스템은, 사진영상 속의 특정 인물별로, 상기 사진영상과 상기 특정인물의 얼굴 특징 및 이름이 저장되는 등록 데이터베이스; 사진에 대해 얼굴특징을 추출하고, 추출된 얼굴이 있는 사진들을 상기 등록 데이터베이스에 등록하는 등록부; 등록되지 않은 새로운 사진이 입력되면 사진 내 인물의 이름을 추정하는 질의부; 및 검색할 인물 이름이 들어오면, 이름이 부여된 얼굴들 중에서 검색할 이름이 들어간 얼굴을 검색하고, 해당 얼굴이 들어간 사진을 검색하고, 상기 사진들을 보여주는 검색부를 포함하는 것을 특징으로 한다. 본 발명에 의하면, 다량의 포토 앨범(Photo Album) 사진들에 대해서, 몇 개의 사진들에 대해서만 수동으로 등록을 하면 나머지 사진들에 대하여 신속하게 등록할 수 있다. 또한, 등록된 사진들을 인물의 이름으로 신속하게 검색할 수 있으므로, 다량의 사진들을 효율적으로 관리할 수 있다.
Abstract:
PURPOSE: An apparatus and a method for extracting user's 3D facial expression are provided to improve the stability about the 3D expression information. CONSTITUTION: A face image of a user is received(212). In the face image, an expression is extracted. 3D expression information is generated(214). 3D expression model with 3D expression information is created(216). Muscle control points are added to the 3D expression model. The 3D expression model is reorganized(218). The reorganized 3D expression model is reflected to the user's avatar(220).
Abstract:
PURPOSE: A camera and a photographing guide information providing method are provided to guide other persons so as to photograph a subject with a background and composition desired by a user in case the user has not a tripod. CONSTITUTION: A camera comprises a photographing unit(120), a display unit(130), a user interface(140), a storage unit(150), and a controller(110). The photographing unit obtains an image by photographing a subject. The display unit displays the obtained image. The user interface receives a user input. The storage unit stores the obtained image. The controller extracts the first scene information which shows the feature of the first scene if the user input for requesting the scene setting is received and extracts the second scene information which shows the feature of the second scene which is captured in the standby state of a photographic mode.
Abstract:
A device and a method for determining face view, and a face detection device and method applying the same are provided to correctly detect a face included in an image at high speed simultaneously with a view class by determining whether a sub window image corresponds to the face and estimating/verifying the view for the sub window image corresponding to the face. A view estimator(210) estimates at least one view class for a current face corresponding to a face. An independent view verifier(230) determines the final view class of the current image by verifying at least one estimated view class independently. The view estimator is implemented by connecting a plurality of levels in a cascade structure. An upper level comprises entire or partial view set, and a lower level comprises an individual view class. The estimator estimates at least one partial view set from entire view set and estimates at least one individual view class from at least one extracted partial view set. The independent view class verifier includes a plurality of view class verifiers. Each view class verifier is implemented by connecting each stage including a plurality of classifiers in the cascade structure.
Abstract:
A video image tracking method and an apparatus thereof are provided to add a new target and remove an existing target easily, and implement embedded software or a chip by reducing calculation and a memory amount required for capturing a target at various angles. A tracking unit(10) determines a target candidate according to each frame by tracking a target model. A detecting unit(20) detects a target image at certain frame intervals. A controller(30) updates the target model by using the target candidate determined by the tracking unit and the target image detected by the detecting unit, and initializes tracking. A tracking position determining unit(11) determines the target candidate in a frame desired for tracking in consideration of statistical distribution characteristics of the target candidate. A histogram extracting unit(12) extracts a histogram with respect to the statistical distribution characteristics of the target candidate determined by the tracking position determining unit.
Abstract:
Weak classifier들의 cascade와 이전 stage의 결과를 결합하는 구조로 되어 있는 얼굴 검출 방법을 개시한다. Weak classifier는 각각의 haar feature를 정확하고 효과적으로 평가할 수 있도록 modified double sigmoid function에 기반하고 있다. 본 발명의 얼굴 검출 방법은 새로운 weak classifier의 파라미터를 학습하는 방법도 포함한다. face detection, 얼굴 후보 검출, strong classifier, weak classifier
Abstract:
인물 DB 등록사진 수 축소 방법 및 장치, 포토 앨범 시스템의 반자동등록 방법 및 장치와, 이를 이용한 포토앨범 시스템이 개시된다. 그 포토앨범 시스템은, 사진영상 속의 특정 인물별로, 상기 사진영상과 상기 특정인물의 얼굴 특징 및 이름이 저장되는 등록 데이터베이스; 사진에 대해 얼굴특징을 추출하고, 추출된 얼굴이 있는 사진들을 상기 등록 데이터베이스에 등록하는 등록부; 등록되지 않은 새로운 사진이 입력되면 사진 내 인물의 이름을 추정하는 질의부; 및 검색할 인물 이름이 들어오면, 이름이 부여된 얼굴들 중에서 검색할 이름이 들어간 얼굴을 검색하고, 해당 얼굴이 들어간 사진을 검색하고, 상기 사진들을 보여주는 검색부를 포함하는 것을 특징으로 한다. 본 발명에 의하면, 다량의 포토 앨범(Photo Album) 사진들에 대해서, 몇 개의 사진들에 대해서만 수동으로 등록을 하면 나머지 사진들에 대하여 신속하게 등록할 수 있다. 또한, 등록된 사진들을 인물의 이름으로 신속하게 검색할 수 있으므로, 다량의 사진들을 효율적으로 관리할 수 있다.