Abstract:
본 발명은 영상 기반의 적응적인 특징점 선택 및 선별적 예상 경로 추정에 기반한 비정형 물체 추적 방법으로서, (a) 물체를 감지하는 단계; (b) 상기 감지된 물체에 기초하여 물체 영역과 주변 영역을 구분하여 설정하는 단계; (c) 상기 설정된 물체 영역과 주변 영역 각각으로부터 특징점을 추출하는 단계; (d) 상기 추출된 물체 특징점과 주변 영역 특징점에 기초하여 물체 특징점의 분별력을 판단하는 단계; (e) 상기 추출 및 갱신된 물체의 특징점에 기초하여 상기 주변 영역의 변화맵을 생성 및 갱신하는 단계; (f) 상기 주변 영역의 특징점 변화 정도에 기초하여 물체가 이동 가능한 영역을 한정 추출하는 단계; 및 (g) 상기 분별력의 판단 결과 분별력이 있는 물체 특징점을 적응적으로 선택하여 상기 한정 추출된 물체 이동 가능 영역 내에서 물체를 추적하는 단계를 포함하여 이루어진다.
Abstract:
According to the present invention, provided is an apparatus for controlling artificial prosthetic hands, comprising: a head-up display interface unit; and a control unit for controlling artificial prosthetic hands in response to the inputs of the head-up display interface unit, wherein the apparatus for controlling artificial prosthetic hands receives inputs by using gazes or eye movements of a user.
Abstract:
The present invention relates to a method and a device for detecting an abnormality of a line which is very useful in processing an image ing real time, in multiple channels on a line such as a power line, and a system using the same. The method for detecting abnormality of a line in accordance with an embodiment of the present invention includes a step of obtaining a line image by focusing on the line with a camera; a step of filtering the obtained image to extract energy corresponding to a predetermined frequency area and a predetermined direction; and a step of determining the abnormality of the line when the energy of the predetermined direction and the predetermined frequency according to the filtering is more than a preset reference value.
Abstract:
A sociability training device and a method thereof are disclosed. The sociability training device which includes an input unit receives the behavior information of a trainee; a trainee state analysis unit which distinguishes the behavior information of the trainee into one among excitement information, pleasure information, and the closeness information and analyzes the state of the trainee; a training mode determining unit which determines a training mode according to the results of the analysis of the trainee; and a robot control unit which controls a robot according to the determined training mode. [Reference numerals] (AA) Start;(BB) End;(S401) Receive various pieces of input information;(S402) Analyze the state of a trainee;(S403) Select a training mode;(S404) Determine the speed of a robot and the robot's action range;(S405) Control the robot
Abstract:
A driving roller for a cable testing robot for passing obstacles which tests the state of a cable while moving in the longitudinal direction of a transmission line and a cable testing robot including the same are provided. The driving roller includes a cylindrical body; and a first roller groove and a second roller groove formed along the lateral periphery of the body. The second roller groove is formed on the outer side of a radial direction of the body in comparison with the first roller groove. An opening of the first roller groove is connected to the second roller groove, and the cross sectional area of the first roller groove is smaller than that of the second roller groove so that the transmission line is received in the first roller groove, and an obstacle which is formed on the middle part of the transmission line and of which the cross sectional area is larger than that of the transmission line is received in the second roller groove.
Abstract:
사회성 훈련 장치 및 그 방법이 개시된다. 피훈련자의 행동 정보를 입력받는 입력부; 입력받은 피훈련자의 행동 정보를 분석하여 피훈련자의 상태를 분석하는피훈련자 상태 분석부; 피훈련자의 분석 상태에 따라 훈련 모드를 결정하는 훈련 모드 결정부; 결정된 훈련 모드에 따라 로봇을 제어하는 로봇 제어부를 포함하는 사회성 훈련 장치가 개시된다. 또한 피훈련자의 행동 정보를 입력받는 단계; 입력받은 피훈련자의 행동 정보를 분석하여 피훈련자의 상태를 분석하는 단계; 및 피훈련자의 분석 상태에 따라 훈련 모드를 결정하는 단계; 결정된 훈련 모드에 따라 로봇을 제어하는 단계를 포함하는 사회성 훈련 방법이 개시된다.
Abstract:
PURPOSE: A robot-based autism diagnostic device that uses an EEG(electroencephalogram), and a method thereof are provided to use an EEG characteristic of an autistic child which is differently observed from the EEG characteristic of a normal child under a specific event circumstance which is generated by a robot, thereby performing an early diagnosis which is required for an early treatment. CONSTITUTION: A robot-based autism diagnostic device that uses an EEG includes an EEG signal detector(100) that detects an EEG signal which is generated by an event which is provided by the EEG signal detector which is attached to the head of a child with an autistic symptoms; and a robot(200) that diagnoses the progress of autism by receiving the EEG signal which is detected by the EEG signal detector(100).
Abstract:
PURPOSE: A system and a method for sensing the face of an intruder are provided to photograph every portion of an intruder with a PTZ(Pan, Tilt, Zoom) camera and to reconstruct even concealed portion of the intruder by combination of every photographed image, thereby correctly detecting the face of the intruder. CONSTITUTION: An intruder detecting unit(300) detects an intruder by receiving images photographed by a PTZ camera(100). An intruder tracking unit(400) tracks the detected intruder. A face sensing unit(500) senses the face of the intruder from the images. A control unit(200) determines the degree of discrimination of the detected face. When it is not possible to discriminate the face of the intruder, a face combination generator(600) senses the face of the intruder using the combination of the images. [Reference numerals] (100) PTZ camera; (200) Control unit; (300) Intruder detecting unit; (400) Intruder tracking unit; (500) Face sensing unit; (600) Face combination generator; (700) Image transmission unit; (800) Memory unit; (810) Pan mode photographed-image DB; (820) Tilt mode photographed-image DB; (830) Zoom mode photographed-image DB
Abstract:
PURPOSE: A device and a method for training sociality are provided to actively control a robot according to the behaviors of autistic children and to train the sociality of the autistic children. CONSTITUTION: A device for training sociality comprises an input unit, a trainee condition analyzing unit, a training mode decision unit, and a robot control unit. The input unit receives behavior information of the trainee(S401). The trainee condition analyzing unit analyzes conditions of the trainee by analyzing the behavior information of the trainee(S402). The training mode decision unit decides the training mode by analyzing the conditions of the trainee(S403). The robot control unit controls the robot according to the training mode(S405). [Reference numerals] (AA) Start; (BB) End; (S401) Receive various input information; (S402) Analyze conditions of a trainee; (S403) Determine a training mode; (S404) Determine the speed and behavior range of a robot; (S405) Control the robot
Abstract:
본 발명은 물체의 시점 변화에 강건한 윤곽선 기반의 범주 물체 인식 방법 및 장치에 관한 것으로, 보다 상세하게는 영상데이터를 입력받으면 상기 영상데이터의 윤곽선을 추출하고, 상기 윤곽선에서 특징점들을 추출하여 물체를 바라보는 시점을 미리 가정하지 않고 특징점 쌍에 대해 물체의 시점 변화에 강건한 형태기술자 벡터를 이용하여 범주 물체를 인식하는, 물체의 시점 변화에 강건한 윤곽선 기반의 범주 물체 인식 방법 및 장치에 관한 것이다. 이를 위하여, 본 발명에 따른 물체의 시점 변화에 강건한 윤곽선 기반의 범주 물체 인식 방법은 (a) 영상데이터를 입력받는 단계와 (b) 상기 영상데이터 중에서 범주 물체를 인식하는 단계 및 (c) 범주 물체 인식 결과를 출력하는 단계를 포함한다. 이에 따라, 물체의 시점에 대한 가정 없이도 다양한 시점에서의 범주 물체를 인식할 수 있다.