http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
엄경배,박기홍 群山大學校自然科學硏究所 1990 自然科學硏究 Vol.5 No.-
The Binary Decision(B.D.) method can evaluate any switching function in a number of steps not exceeding the number of input variables. It is shown that the simplification for minimal trees can be achieved through the state reduction methods to obtain decision algorithm. An algorithm for the optimal B.D. tree was systematically developed by using the incomplete multiple output function and the residual characteristic function.
형태학적 형태 분해 방법을 이용한 물체 인식에 관한 연구
嚴景培,金準哲,李俊煥 全北大學校 1994 論文集 Vol.37 No.-
Mathematical morphology based on the set theory has been applied to various areas in image processing. In this paper, we propose a new method for object recognition based on morphological shape decomposition. The first step of our method decomposes a binary shape into a union of simple binary shapes, and then a new tree structure is constructed which can represent simple binary shapes of objects. Finally, we obtain the feature informations from the tree structures and calculate matching scores using efficient matching measure. The experimental results of the proposed method show the good recognition rate.
혼합된 칼라 잡음 제거를 위한 조건적인 퍼지 클러스터 필터
엄경배,한서원 群山大學校基礎科學硏究所 1998 基礎科學硏究 Vol.13 No.-
Color image is more powerful than gray one in human visual perception. Therefore, color image processing becomes important area. Color images are often corrupted by noises due to the input sensor or channel transmission errors. Some filtering techniques such as vector median, mean filter, and vector α-trimmed mean filter have been used for color noise removal. Among them, vector α-trimmed mean filter gave the best performance in the mixed color noise. But, there are edge shift and blurring effect, because vector α-trimmed mean filter is uniformly processed across the image. So, we proposed the conditional fuzzy cluster filter to improve this problems. Simulation results showed that the proposed scheme improves the NCD measure and visual quality over the conventional vector α-trimmed mean filter in the mixed color noise.
형상학을 이용한 골격과 모서리 추출 및 영상 압축에 관한 연구
엄경배,박기홍 群山大學校基礎科學硏究所 1992 基礎科學硏究 Vol.7 No.-
In this paper, we present a image compression method based on morphological skeleton and edge. A compressed data is restored without any loss information. A new method results in improved computational performance and reduced memory storage requirements. If the size of the maximal element of a skeleton point is larger than the size of unit structuring elements, the new method will result in at least a 25% computational savings. Computational efficiency increases nonlinearly as the size of the maximal element increases.
Mathematical Morphology를 이용한 영상 데이타 압축에 관한 연구
엄경배,이양원,신정훈 群山大學校 1991 論文集 Vol.18 No.-
Digital image compression is needed more and more in data communication for various purpose. A new image compression method, based on morphological transformation, is presented in this paper, The object in an image can be compressed without any loss information by an iterative erosion operation. The original image is restored by the iterative dilation operation with the same structure element. God result is obtained in binary image data compression. The important characteristic of this method is that the compressed data may be used directly to analysis an image.
퍼지 시스템에서 inverse problem의 해에 대한 알고리즘
엄경배,박인철 群山大學校 1991 論文集 Vol.18 No.-
퍼지 시스템에 대한 inverse problem 은 E.Sanchez, C.P.Pappis와 M.Sugeno [2,5,6]에 의해 연구되어졌다. 이 논문에서는 ??인 a가 존재할 필요충분조건을 증명하고 [Theorem 3], 이 조건을 이용하여 알고리즘을 구성하였다. 또한, PASCAL언어로 구현함으로써 예제(1)과 예제(2)의 상한과 하한을 얻었다. 이 문제에서 상한과 하한의 생성은 전문가 시스템이나 추론 기법 등에서 유용하게 이용될 수 있다.
멀티 프로세서에서 타스크 할당 및 시스템 구현에 관한 연구
박기홍,엄경배 군산대학교 1988 論文集 Vol.15 No.-
In this paper we make up a new compiler of having the level of subset which creat the object codes in multiprocessor. The multiprocessor is composed of the time shared bus structure of being three PES board, common memory board, system board and SIB board. We input the sub-basical source program into a composed compiler and extract all the possible parallelism in a job.In other to change sequential program into suitable parallel form in multiprocessor, we partition a job into several tasks which can perform parallel processor. We make a practice divided tasks simultaneously in each diffrent processor elements on the system of tightly coupled multiprocessor.
Relevance feedback을 이용한 감성언어 기반 영상 검색
박중수,엄경배 국립7개대학공동논문집간행위원회 2002 공업기술연구 Vol.2 No.-
We propose an on-line query mechanism for emotion-based similarity retrieval of images in database. It uses emotional languages for the first search. We know that it successfully catches the user's emotion. For the second search, we use relevance feedback algorithm to reflect the gap between the user's emotion and computer calculated contents. Relevance feedback successfully analyzes the data that is user's estimation of images and reduces the gap between user and computer. The features and similarity measures are adopted from MPEG-7 color descriptors which are proper retrieval of large multimedia databases. We use 286 images for experiment which are corrected from internet. The result shows that the system is successful.
DSP56001을 이용한 물체의 불변 인식 시스템의 구현
한서원,엄경배 群山大學校基礎科學硏究所 1998 基礎科學硏究 Vol.13 No.-
In this study, an image processing board was developed and an invariant object recognition system was implemented on the board. The board has developed for general purpose using DSP56001. The object recognition system based on Fourier descriptor is invariant to similarity transformations such as translation, rotation and scaling. For the classification, the system adopts K-NN(K-Nearest Neighbor) algorithm. The performance of the system is evaluated with non-overlapped 2-dimensional automobile objects. The system showed perfect classification for the transformed objects. For the recognition, it takes about 0.664 second including appropriate image processing.
Boosting algorithm을 이용한 다중 질의 영상검색
박중수,엄경배 국립7개대학공동논문집간행위원회 2002 공업기술연구 Vol.2 No.-
Content-based image retrieval uses extracted image contents. But if the contents are consist of many features then the retrieval process may speed down. In this paper, we use boosting algorithm to reduce the redundancy of features. We experiment 286 wallpapers and get the successful result.