http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
민성재(Sung-Jae Min),이수영(Soo-Yeong Yi) 제어로봇시스템학회 2018 제어·로봇·시스템학회 논문지 Vol.24 No.9
Non-frontal cameras have an image sensor that is intentionally slanted compared to the lens. Because the sensor and the lens are not parallel, the sharp focus plane of the camera is also slanted with variable object distance on the plane according to the lens’s law. In this study, we propose a high-definition image-acquisition method that uses a non-frontal camera. A series of images are obtained by rotating the non-frontal camera about an optical axis, and each of the images has a sharp focus plane of variable object distance. We can compose a sharp focus image by selecting focused pixels among the series of images according to a focus measure. The initial results of the non-frontal camera and further work to compose high-definition images are presented in this study.
효율적인 꼭짓점 검출 및 2 차원 LIDAR 데이터 정합 알고리즘
민성재(Sung-Jae Min),이수영(Soo-Yeong Yi) 제어로봇시스템학회 2021 제어·로봇·시스템학회 논문지 Vol.27 No.1
The feature detection of LIDAR data and their matching are essential for the localization and map-making of a mobile vehicle. In this study, an efficient detection algorithm for the corner points in a LIDAR data is proposed using line segments or corner points as general features. The proposed algorithm detects the corner points using the relationship between the adjacent measurement points. A virtual line segment consisting of two corner points detected in a data set is used by the matching algorithm in this study to compute the transformation including the translation and the orientation between two measurement data sets. Because the length of a line segment in a data set is preserved by the transformation, finding a corresponding line segment with similar length in the other measurement data set is easy. Experimental results verify the performance of the corner detection and the matching algorithm proposed in this study.
적응적 학습 파라미터의 고정점 알고리즘에 의한 독립성분분석의 성능개선
조용현,민성재,Cho, Yong-Hyun,Min, Seong-Jae 한국정보처리학회 2003 정보처리학회논문지B Vol.10 No.4
This paper proposes an efficient fixed-point (FP) algorithm for improving performances of the independent component analysis (ICA) based on neural networks. The proposed algorithm is the FP algorithm based on Newton method for ICA using the adaptive learning parameters. The purpose of this algorithm is to improve the separation speed and performance by using the learning parameters in Newton method, which is based on the first order differential computation of entropy optimization function. The learning rate and the moment are adaptively adjusted according to an updating state of inverse mixing matrix. The proposed algorithm has been applied to the fingerprints and the images generated by random mixing matrix in the 8 fingerprints of 256${\times}$256-pixel and the 10 images of 512$\times$512-pixel, respectively. The simulation results show that the proposed algorithm has the separation speed and performance better than those using the conventional FP algorithm based on Newton method. Especially, the proposed algorithm gives relatively larger improvement degree as the problem size increases. 본 연구에서는 뉴우턴법의 고정점 알고리즘에 적응 조정이 가능한 학습 파라미터를 이용한 효율적인 신경망 기반 독립성분분석기법을 제안하였다. 이는 엔트로피 최적화 함수의 1차 미분을 이용하는 뉴우턴법의 고정점 알고리즘에서 학습율과 모멘트를 역혼합행렬의 경신 상태에 따나 적응조정되도록 함으로써 분리속도와 분리성능을 개선시키기 위함이다 제안된 기법을 256$\times$256 픽셀의 8개 지문과 512$\times$512 픽셀의 10개 영상으로부터 임의의 혼합행렬에 따라 발생되는 지문과 영상의 분리에 적용한 결과, 기존의 고정점 알고리즘에 의한 결과보다 우수한 분리성능과 빠른 분리속도가 있음을 확인하였다. 특히 제안된 알고리즘은 문제의 규모가 클수록 분리성능과 분리속도의 개선 정도가 큼을 확인하였다.