Abstract:
PURPOSE: A power quality classifying method using a voice processing technique, a power quality monitoring device thereof, and a power quality monitoring system thereof are provided to accurately classify an external disturbance signal. CONSTITUTION: A power quality monitoring device (100) includes a transformer (110), an A/D converter (120), a classification processor (130), and a memory (140). The transformer reduces a power signal transmitted from a power system to a reference level. The A/D converter converts the power signal into a digital signal. A classification processor classifies power quality of the power signal by using a mel frequency cepstral coefficient (MFCC) value about the power signal. The memory stores data used for power quality classification in the classification processor. [Reference numerals] (110) Transformer; (120) A/D converter; (130) Classification processor; (131) Power signal statistics unit; (132) Initial classification unit; (133) MFCC value sampling unit; (134) GMM classification unit; (140) Memory
Abstract:
본 발명은 차영상 엔트로피를 이용한 시선 추적 장치 및 그 방법에 관한 것으로, 촬영장치를 통하여 촬영된 영상 중 눈 영역 영상을 검출하기 위한 영상 검출부, 눈 영역 영상과 다수의 시선 위치별 레퍼런스 영상의 차영상 엔트로피를 연산하고 레퍼런스 영상을 선택하는 영상 선택부 및 선택된 레퍼런스 영상에 따른 눈동자의 응시방향을 인식하는 시선 추적부를 포함하는 시선 추적 장치 및 이에 따른 시선 추적 방법을 제공함으로써 손가락, 음성, 마우스를 대신하여 컴퓨터, 휴대단말기 및 전자 기기의 제어가 가능하게 된다. 시선 추적, 차영상, 엔트로피
Abstract:
본 발명에 따른 이파리 영상 자동인식을 이용한 식물 분류 방법은 입력 이파리 영상이 영상장치를 통해 획득되는 단계, 입력 이파리 영상에서 이파리 중심점으로부터 이파리 외곽선까지의 거리를 이용하여 거리 파라미터가 추출되는 단계, 입력 이파리 영상에서 이파리의 형태적 특징을 이용하여 형태 파라미터가 추출되는 단계 및 거리 파라미터 및 형태 파라미터를 기준으로 인식 후보 이파리 영상과 비교하여 입력 이파리 영상의 식물 분류가 결정되는 단계를 포함한다. 거리 파라미터가 추출되는 단계는 입력 이파리 영상에서 이파리 영역이 검출되는 단계, 이파리 영역에서 이파리 중심점 및 이파리 외곽선이 검출되는 단계, 이파리 중심점으로부터 이파리 외곽선을 구성하는 모든 픽셀들까지의 거리가 계산되는 단계 및 거리를 기준으로 거리 파라미터가 추출되는 단계를 포함한다.
Abstract:
A plant classification method using automatic leaf image recognition according to the present invention comprises the following steps: obtaining a leaf image by an imaging device; extracting a distance parameter using the distance between the central point and the outline of the leaf in the leaf image; extracting a shape parameter using morphological features of the leaf in the leaf image; and determining the plant classification of the inputted leaf image by comparing the leaf image with recognition candidate leaf images based on the distance parameter and shape parameter. The distance parameter extraction step includes the following steps: detecting a leaf area in the leaf image; detecting the central point and the outline of the leaf in the leaf area; calculating distances from the central point of the leaf to all the pixels on the outline of the leaf; and extracting a distance parameter based on the calculated distances. [Reference numerals] (AA) Start;(BB) Input image;(CC) Detect a leaf area;(DD) Extract the outline and the central point of the leaf;(EE) Calculate the distance between the outline and the central point;(FF) Extract a feature parameter based on the distance;(GG) Extract a feature parameter based on morphological and geographical features;(HH) Recognize the plant (18 steps) using the leaf features;(II) End
Abstract:
A method and an apparatus for detecting a signal and a recording medium on which a program for executing the method is recorded are provided to detect a desired signal in noise environment by using a delta spectrum entropy value. A divider(130) divides a received input signal by a frame unit. A converter(140) converts input signals existing in the first and second frames into frequency signals. A production unit(150) produces the first power spectrum information and the second power spectrum information by using the converted frequency signals. An obtaining unit(160) obtains a delta spectrum entropy value corresponding to a difference of the produced power spectrum information. A determining unit(170) compares the delta spectrum entropy value with a threshold value.
Abstract:
Disclosed are an apparatus and a method for detecting a vein and extracting vein features from a leaf image. The apparatus and the method comprises the following steps of: inputting a leaf image using a photographing device of a computer or a mobile terminal; detecting a vein area in the leaf image by the processor of the computer or the mobile terminal; detecting, by the processor, a vein in the vein area and obtaining an vein image of the detected vein; detecting, by the processor, the main vein of the vein by horizontally projecting the vein image; determining, by the processor, the direction of the vein by perpendicularly projecting the vein image; and thinning the vein and extracting the features of the vein by the processor. As a result of the automatic processes from image input to leaf feature extraction, the apparatus and the method have efficient and high leaf recognition performance. [Reference numerals] (110) Input an image;(120) Detect a vein area;(130) Obtain a vein image;(140) Detect a main verin;(150) Determine the direction of the vein;(160) Thin the vein and extract the features of the vein;(AA,BB) Start
Abstract:
PURPOSE: A gaze tracking apparatus and method using difference image entropy is provided to control the various kinds of menu by using a computer, portable terminal, and camera mounted in the electronic device instead of a finger, voice, mouse by recognizing/tracing the pupil eye-gaze direction. CONSTITUTION: The gaze tracking apparatus using the difference image entropy(100) includes an image detection part, image selection part, and gaze tracker. The image detection part(120) detects the eye area image from the image taken by the camera(10). The image selection part(130) compares the eye area image with the reference images according to a plurality of eye positions and calculates the difference image entropy. The image selection part selects the reference image showing the smallest value of the difference image entropy. The gaze tracker(140) recognizes the eye-gaze direction of the pupil according to the selected reference image.
Abstract:
The disclosed technology relates to a power signal generating method. The virtual power signal generating method randomly generating a power signal which may be generated in a distribution system comprises a step of inputting an irregular first power signal; a step of classifying the inputted first power signal to a type including at least one among an instantaneous voltage drop state, a harmonic state, a normal voltage state, or a transient state according to methods of evaluating power quality; a step of dividing the classified first power signal into one or more segments; a step of generating a successive second power signal using some of the segments; and a step of smoothing the second power signal. Therefore, the power signal generating method is performed to effectively analyze the power quality. [Reference numerals] (110) Inputting a first power signal;(120) Classifying the inputted first power signal to a type;(130) Dividing the classified first power signal into segments;(140) Generating a second power signal;(150) Smoothing;(AA) Start;(BB) End
Abstract:
본 발명의 일 측면에 따르면 신호 검출 방법이 개시된다. 본 발명의 일 실시예에 따른 신호 검출 방법은 수신된 입력 신호를 일정한 프레임 단위로 분할하고, 제1 프레임 및 제2 프레임에 존재하는 각각의 입력 신호를 주파수 신호로 변환하고, 변환된 주파수 신호로부터 제1 파워 스펙트럼 정보 및 제2 파워 스펙트럼 정보를 산출한 후, 산출된 파워 스펙트럼 정보의 차이에 상응하는 델타 스펙트럼 엔트로피값을 획득하는 델타 스펙트럼 엔트로피값을 획득하여, 델타 스펙트럼 엔트로피값과 임계값을 비교하여, 수신된 입력 신호 중 임의의 프레임에 임의의 입력 신호가 포함되는지 여부를 판단한다. 본 발명에 의하면 델타 스펙트럼 엔트로피값을 이용하여 잡음 신호가 존재하는 잡음 환경에서도 원하는 신호의 검출이 가능한 장점이 있다. 델타 스펙트럼, 델타 스펙트럼 엔트로피, 신호 검출