• 한국농업기계학회:학술대회논문집
• /
• 한국농업기계학회 1996년도 International Conference on Agricultural Machinery Engineering Proceedings
• /
• pp.773-779
• /
• 1996

Dried oak mushroom have complex and various visual features. Grading and sorting of dried oak mushrooms has been done by the human expert. Though actions involved in human grading looked simple, a decision making underneath the simple action comes from the result of the complex neural processing of the visual image. Through processing details involved in human visual recognition has not been fully investigated yet, it might say human can recognize objects via one of three ways such as extracting specific features or just image itself without extracting those features or in a combined manner. In most cases, extracting some special quantitative features from the camera image requires complex algorithms and processing of the gray level image requires the heavy computing load. This fact can be worse especially in dealing with nonuniform, irregular and fuzzy shaped agricultural products, resulting in poor performance because of the sensitiveness to the crisp criteria or specific ules set up by algorithms. Also restriction of the real time processing often forces to use binary segmentation but in that case some important information of the object can be lost. In this paper, the neuro net based real time recognition algorithm was proposed without extracting any visual feature but using only the directly captured raw gray images. Specially formated adaptable size of grids was proposed for the network input. The compensation of illumination was also done to accomodate the variable lighting environment. The proposed grading scheme showed very successful results.

• ETRI Journal
• /
• 제38권6호
• /
• pp.1042-1051
• /
• 2016

The mobile hotspot network (MHN), which is capable of providing a data rate of gigabits per second at high speed, is considered a potential use case of the future enhanced mobile broadband for 5G. Because a unidirectional network deployment has been considered for an MHN, non-orthogonal multiple access (NOMA) can be employed to improve the system performance. For a practical implementation of NOMA under an MHN highway scenario where multiple pieces of MHN terminal equipment are served through the same beam simultaneously, a NOMA transceiver is proposed in this paper. For the NOMA transmitter, Gray-coded QAM constellation mapping is extended to arbitrary modulation order q. For the NOMA receiver, successive interference cancellation (SIC) is no longer necessary, and instead, a parallel demodulation is proposed. The numerical and simulation results suggest that the proposed NOMA transceiver outperforms the conventional NOMA SIC receiver and can be flexibly used for an MHN highway scenario.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.644-647
• /
• 2003

This paper describes a neural network based machine vision system designed for inspecting yellow beans in real time. The system consists of a camera. lights, a belt conveyor, air ejectors, and a computer. Beans are conveyed in four lines on a belt and their images are taken by a monochrome line scan camera when they fall down from the belt. Beans are separated easily from their background on images by back-lighting. After analyzing the image, a decision is made by a multilayer artificial neural network (ANN) trained by the error back-propagation (EBP) algorithm. We use the global mean, variance and local change of gray levels of a bean for the input nodes of the network. In an our experiment, the system designed could process about 520kg/hour.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2006년도 춘계학술대회 논문집
• /
• pp.61-62
• /
• 2006

A new neural network classifier is proposed for the automatic real-time surface inspection of high-speed cold steel strips having 11 different types of defects. 46 geometrical and gray-level features are extracted for the defect classification. 3241 samples of Posco's Kwangyang steel factory are used for training and testing the neural network classifier. The developed classifier produces plausible 15% error rate which is much better than 20-30% error rate of human vision inspection adopted in most of domestic steel factories.

• 한국조경학회지
• /
• 제37권5호
• /
• pp.53-63
• /
• 2009

본 연구는 공원녹지체계를 광의의 개념으로서 오픈스페이스적으로 접근하는 것이라고 설정하고, 그 방법론을 모색하고 실증적으로 대전광역시 도심지역에 적용하였으며, 그 적용 결과에 기초하여 대전광역시 오픈스페이스 네트워크 구축방안을 제시하였다. 연구방법으로 첫째, 대전광역시 구도심을 대상으로 오픈스페이스 네트워크의 적용 가능성을 검토하며, 둘째, 사례지역의 결과를 토대로 대전광역시의 오픈스페이스 유형별 네트워크를 분석하고, 이를 종합함으로써 대전광역시 오픈스페이스네트워크를 구상토록 하였다. 연구결과로서, 활동오픈스페이스 요소인 관공서 및 업무시설들은 자체적으로 또는 건물 전면부에 공개공지가 조성되어 있고, 이들 전면부 공개공지는 보행로(그레이 오픈스페이스)와 연결되어 있으며, 이들 보행로는 결과적으로 대규모 공원이나 하천(그린 및 블루 오픈스페이스)으로 연결되어, 통합하여 하나의 공간체계를 형성한다고 할 수 있다. 따라서 도시 중심부에는 그동안 공원녹지체계의 기본적 요소라고 할 수 있는 공원과 녹지가 상대적으로 적기 때문에, 공원녹지의 추가적 확보가 이루어져야 공원녹지체계의 개선이 진행된다고 볼 수 있다. 도심부에 새로운 공원녹지의 신규조성은 매우 어렵다는 전제하에 도심부에 마치 실핏줄처럼 분포하고 있는 오픈스페이스 네트워크체계로 접근할 경우, 도심부에 보다 건강한 휴식공간체계가 형성될 것이다.

• 한국정보통신학회논문지
• /
• 제11권6호
• /
• pp.1162-1169
• /
• 2007

본 논문에서는 항만에서 취급하는 컨테이너의 식별자를 인식하는 방법을 제안한다. 실제 컨테이너 영상을 그레이 영상으로 변환한 후, 프리윗 마스크(Prewitt mask)를 적용하여 윤곽선을 검출하고 컨테이너를 식별할 수 있는 개별 식별자의 형태학적 특징 정보를 이용하여 식별자 후보 영역을 추출한다. 검출된 식별자 후보 영역은 개별 식별자 영역외에 잡음 영역이 포함되어 있으므로 4방향 윤곽선 추적 알고리즘과 Grassfire 알고리즘을 적용하여 잡음을 제거하고 개별 식별자들을 각각 객체화한다. 잡음이 제거된 식별자 후보 영역에서 객체화 한 개별식자는 컨테이너 식별을 위해 FCM 기반 퍼지 RBF 네트워크를 적용하여 인식한다. 본 논문에서 제안한 컨테이너 식별자 인식 방법의 성능을 평가하기 위해 실제 컨테이너 영상 300장을 대상으로 실험한 결과, 기존의 방법보다 인식 성능이 개선되었음을 확인할 수 있었다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제4권1호
• /
• pp.56-62
• /
• 2004

This research attempts to suggest a detecting method for cutting position of an object using the neural network, which is one of intellectual methods, and the digital image processing method. The extraction method of object information using the image data obtained from the CCD camera as a replacement of traditional analog sensor thanks to the development of digital image processing. Accordingly, this research determines the threshold value in binary-coding of an input image with the help of image processing method and the neural network for the real-time gray-leveled input image in substitution for lighting; as a result, a specific position is detected from the processed binary-coded image and an actual system designed is suggested as an example.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.864-869
• /
• 1993

This paper presents the problem of automatically recognizing embossed or molded characters which are raised from the side wall on rubber tire. In the tire image objects have approximately the same gray-value as the background and because of the tire geometry, illumination of the surface is nonhomogenous. Therefore it is difficult to recognize the raised tire character. In this paper, for describing the process of processing three steps have been proposed: 1) MIN & MAX smoothing operation filter, 2) edge detection algorithm using modified laplacian operator, 3) noise rejection. Afterwards, segmentation is done to attain characters from tire image by projection method. The recognition of the characters in the segmented image is carried out by using multilayered neural network, which is insensitive to the noise.

• 한국생산제조학회지
• /
• 제9권3호
• /
• pp.103-110
• /
• 2000

Shaft encoder which encodes the rotational angle of a shaft becomes more important recently due to factory automation and office automation. Although an absolute type encoder is more dsirable due to its convenience an incremental encoder is commonly used because of its cost and technical difficulties Fabricating a high resolution absolute encoder is very diff-cult because the physical size is limited by currently available technology. In order to overcome this difficulty Moire fringe can be used incorporated with gray code. In order to measure the position of fringes which move as the code disk rotates a neural network was developed in this paper. Formerly fringe position is usually measured by a sophisticated software which needs a little long calculation time. However using nerual network method can eliminate such calculation time even though it needs learning job The pro-posed method is verified through several experiments.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 1999년도 추계학술대회 논문집 - 한국공작기계학회
• /
• pp.320-325
• /
• 1999

An absolute type shaft encoder which utilized moire fringe will be presented in this paper. Linear moire fringe is commonly used to measure the displacement of the linear motion. However, an absolute encoder which measure the rotation angle of a shaft is operated usually with a code disk which the gray code pattern is printed on. Such encoder has inherently resolution limit because of the patterning mechanism and sensing mechanism. In order to measure the position of fringes which move as the code disk rotates, neural network was developed in this paper. Formerly fringe position is usually measured by a sophisticated software, which needs a little long calculation time. However, using neural network method can eliminate such calculation time, even though it needs learning job. The proposed method is verified through several experiments.