아날로그 신경회로망을 이용한 스테레오 정합 ( Stereo Matching Using Analog Neural Network )

도경훈,이준재,조석제,하영호 대한전자공학회 1991 대한전자공학회 학술대회 Vol.1991 No.1

형태학적 세선화 영상의 세그먼트를 이용한 겹쳐진 물체 인식

도경훈 동서대학교 부설 연구소 1996 연구소 논문집 Vol.1 No.-

Mathematical morphology based on set theory is easy to be implemented in parallel and can be applied to various fields in image analysis. Morphological filters can be used effectively for medical and microscopic images to separate overlapping particles. In this paper, separation of overlapping particle image using morphological method is proposed. The extraction of unique marker per every particle and the reconstruction of separated particle from this marker image are done by erosion and conditional dilation, respectively. This approach has the advantages of efficient marker extraction, high separation accuracy, and parallel implementation.

효율적인 온라인 PMSM 매개변수 추정 방법

도경훈,최한호 제어·로봇·시스템학회 2022 제어·로봇·시스템학회 논문지 Vol.28 No.12

. This paper proposes an efficient estimation method for PMSM (Permanent Magnet Synchronous Motor) parameters. We usually need to precisely estimate PMSM parameter values of the flux linkage, stator inductance, and resistance for accurate control, condition monitoring, fault diagnosis and prognosis, or others. The conventional methods to estimate the flux linkage, stator inductance, and resistance involve the manipulation of matrices of size 3 by 3 or higher dimensions, which leads to heavy computational burdens. This paper provides a simpler estimation method involving 2-by-2 matrix manipulations. Using the estimated PMSM parameter values of the flux linkage, stator inductance, and resistance, we design an adaptive PI-type PMSM controller. Through numerical comparative simulations, this paper verifies the practicality and efficiency of the proposed method.

스테레오 정합의 변이 정보를 이용한 영상 분할

도경훈 東西大學校 1996 동서논문집 Vol.2 No.1

본 논문에서는 변이 정보를 이용한 영상 분할 알고리즘을 제안한다. 제안한 영상 분할 방법에서는 스테레오 영상으로부터 영상의 명암도, 경계선 등의 정보와 좌우 영상의 정합으로 얻어지는 변이 정보를 동시에 고려하여 물체의 보다 정확한 분할을 얻는다. 좌우 두 영상은 계층적 구조를 형성하기 위해 웨이브릿 변환에 의해 각 채널별로 분리된 후 스테레오 정합을 통해 변이 정보의 초기치를 얻으며 이로부터 영상은 물체 단위의 분할을 수행하고 그 결과를 이용해서 변이값을 다시 개선하는 과정을 반복한다. 정합이나 분할 과정에서 웨이브릿 변환을 이용한 계층적 구조를 사용함으로써 수행 속도의 개선을 가져왔고 잡음에도 강하다. 실영상과 합성 영상에 대한 실험 결과는 제안한 방법이 기존의 방법에 비해 물체 단위의 분할에 매우 유용하게 사용될 수 있음을 보여준다. In this paper, the algorithm for object segmentation using disparity information is proposed. Left and right stereo images are decomposed into each channel by wavelet filter. And then, disparity value is obtained by hierarcvhical stereo matching. The disparity and segment information is refined iteratively using homogeneous regions segmented by modified blob coloring. The experimental results for various images show that the proposed algorithm has good performance even if the images used in experiments have the unfavorable conditions.

변이간의 적응적 후원을 이용한 이완 스테레오 정합

도경훈,김용숙,하영호 대한전자공학회 1996 전자공학회논문지-B Vol.b33 No.3

This paper presetns an iterative relaxation method for stereo matching using matching probability and compatibility coefficients between disparities. Stereo matching can be considered as the labeling problem of assigning unique matches to feature points of image an relaxation labelin gis an iterative procedure which reduces local ambiguities and achieves global consistency. the relation between disparities is determined from highly reliable matches in initial matching and quantitatively expressed in temrs of compatibility coefficient. The matching results of neighbor pixels support center pixel through compatibility coefficients and update its matching probability. The proposed adaptive method reduces the degradtons on the discontinuities of disparity areas and obtains fast convergence.

살색정보와 움직임 정보를 이용한 영상분석

도경훈,정의윤,조정복 東西大學校 2000 동서논문집 Vol.6 No.-

Objects-based coding algorithms segment image contents into set of objects or regions according to a given model and estimate their parameters like motion, shape, and color. It is known that object-based coding is more efficient coding method in videophone image sequence. It is difficult to detect moving object only using motion information. Chromaticity distribution of skin color is similar region independent of human race. This paper presents efficient image analysis method using both skin color and motion. The proposed algorithm is composed of facial region detection using skin color and region detection using motion information.

유전 알고리즘을 이용한 거리 정보의 추출

김용석,도경훈 東西大學校 1997 동서논문집 Vol.3 No.1

In this paper, a genetic algorithm-based optimization technique for stereo matching is proposed. Stereo matching is the essential process to recover three-dimensional structure of objects. The geometrical difference of left and right images called as disparity is constructed as two-dimensional chromosomes with fitness values inversely proportional to their costs. The cost function is composed of the intensity-difference between two images and smoothness of disparity. The crossover and mutation operators in the two-dimensional chromosomes is described. The operations are affected by the disparities of neighbor pixels. The knowledge-augmented operators is shown to result in rapid convergence and stable result. The genetic algorithm for stereo matching is tested on synthetic and natural images. Experimental results for various images show that the proposed algorithm has good performance even if the image has the unfavorable conditions.

효율적 영역 기반 부호화를 위한 인간 시각 특성에 기반한 계층적 영상 분할

송근원,정광민,도경훈 東西大學校 1998 동서논문집 Vol.4 No.-

본 논문에서는 주관적 화질 및 객관적 화질을 유지하면서 영역 기반 초저속 부호화에서 병목 현상을 유발하는 과다한 윤곽선 정보를 효과적으로 줄일 수 있으며, 또한 각 계층에서의 전송 정보량를 조절할 수 있는 계층적 영상 분할 알고리즘을 제안한다. 제안한 알고리즘은 각 계층에서 수리 형태학에 기반한 계층적 영상 분할과 인간 시각 시스템(HVS)특성을 고려한 영역 병합으로 구성된다. 계층적 영상 분할은 3계층으로 이루어져 있으며, 영상 분할 수 인간 시각이 구분할 수 없는 두 인접 영역을 병합한다. 이때 인간 시각 시스템 특성에 기반하여 병합 순서를 결정하는 병합 우선 순휘 함수(MPF)를 제한하여 영역 병합을 수행한다. 실험을 통하여 제안한 알고리즘은 기존의 알고리즘보다 PSNR은 평균 0.14dB 낮으나, 윤곽선 정보량은 30% 줄일 수 있었으며, 또한 각 계층에서 전송 정보량을 조절할 수 있는 적응성이 있는 영상 분할 구조임을 알 수 있었다. In this paper, a new hierarchical image segmentation algorithm based on human visual system(HVS) characteristic, which can efficiently reduce information quantity to be transmitted without the degradation of the subjective and objective image quality, and which is suitable for region-based very low bit coding because it can control information quantity to be transmitted, is proposed. It consists of image segmentation based on mathematical morphology and region merging considering HVS characteristic for the pairs of two adjacent regions at each level of the hierarchy. Image segmentation is composed of 3-level hierarchical structure and in the region merging structure of each level, we extract the pairs of two adjacent regions which human vision can't discriminate, and then merge them. Since the proposed region merging method selects two neighbor regions to be merged and performs region merging according to merging order based on HVS characteristics, it can efficiently improve bottleneck problem caused by excessive contour information at region-based very low bit rate coding. In experimental results, though the PSNR by the proposed algorithm is about 0.14dB lower than that of conventional method, the subjective image quality is almost the same and the contour information to be transmitted is reduced as much as 23.9% on the average.

소형 유전자 알고리즘을 이용한 스테레오 정합

한규필,도경훈 東西大學校 1999 동서논문집 Vol.5 No.-

Genetic algorithm is an efficient search method using principles of natural selection and population genetic. In conventional genetic algorithms, however, the size of gene pool should be increased to tnsure a convergency. Therefore, many memory spaces and much computation time were needed. Also, sine child chromosomes were generated by chromosome crossover and gene mutation, the algorithms have a complex structure. Thus in this paper, a compact streo matching algorithm using a population-based incremental learning dased on probability vector is proposed to reduce these problems. The PBIL method is modified for matching environment. Since the proposed algorithm uses a probability vwctor and eliminates gene pool, chromosome crossover, and gene mutation, the matching algorithm is simple and the computation load is considerably reduced. Even if the characteristics of images are changed, stable outputs are obtained without the modifification of the matching algorithm.

회전자의 과도한 온도 상승을 줄이는 PI형 PMSM 제어기 설계

도경훈(Gyeong-hun Do),최한호(Han Ho Choi) 제어로봇시스템학회 2022 제어·로봇·시스템학회 논문지 Vol.28 No.2

This study considers the problem of designing a PI-type PMSM (permanent magnet synchronous motor) controller, which can prevent excessive temperature rise in the rotor. The problem was formulated as a nonlinear optimization problem, considering that the flux linkage of a PMSM depends on the rotor temperature value and can be estimated using the voltage input value. This study proposes a design procedure for solving the nonlinear optimization problem using the particle swarm optimization (PSO) algorithm, a metaheuristic search algorithm. Simulation studies using a practical PMSM model verified that the design problem can be successfully solved using a metaheuristic search algorithm.