Vector quantizer를 이용한 near lossless 의학 영상 압축

송영철(Y. C. Song),안창범(C. B. Ahn) 대한전기학회 1996 대한전기학회 학술대회 논문집 Vol.1996 No.7

적응 템플릿 필터링에서의 Edge artifact 제거

안창범(C. B. Ahn),송영철(Y.C. Song) 대한전기학회 2000 대한전기학회 학술대회 논문집 Vol.2000 No.7

삼차원 의학 영상에서 시각적으로 향상된 정합을 사용한 프랙탈 부호화

안창범,송영철,신흥식,Ahn, C.B,Song, Y.C,Shin, H.S 대한의용생체공학회 1996 의공학회지 Vol.17 No.3

A new three dimensional fractal coding is proposed with a perceptually enhanced matching. Since most of medical images (e.g. computed tomoyaphy or magnetic resonance images) have three dimensional characters, searching regions are extended to adjacent slices. For a perceptually enhanced matching, a high frequency boost filter is used for pre-filtering images to be encoded, and a least mean square error matching is applied to the edge enhanced Images rather than the original images. From simulation with magnetic resonance images($255\times255$, 8bits/pixel), reconstructed images by the proposed compression algorithm show much improved subjective image quality with higher peak signnal-to-noise ratio compared to those by existing fractal coding algorithms at compression ratios of about 10.

자기공명영상의 신호대잡음비 향상을 위한 다중 템플릿 적응 필터링

안창범(Ahn C . B .),송영철(Song Y . C .),박대준(Park D . J .) 한국화상학회 1997 한국화상학회지 Vol.3 No.1

본 논문에서는 해상도의 손상 업이 자기공명영상에서의 신호대잡음비를 개선시킬 수 있는 적응 필터링을 제안하였다. 제안한 방법에서는 템플릿 형태가 고정되어 있는 기존의 필터링 방법과 달리 다중 템플릿들을 정의하였으며, 제안된 다중 템플릿들 중에서 각 픽셀에서 주변의 영상 구조와 가장 잘 매칭이 되는 템플릿을 선택하였다. 제안한 알고리즘을 통해 가장 잘 매칭이 되는 템플릿을 선택함으로써 에지 열화를 최소화하였으며, 신호대잡음비 향상을 얻을 수 있었다. 다중 템플릿을 사용하는 제안된 형태 적응 필터의 성능을 기존의 2-D LLSE 필터와 방향 적응 재귀 필터의 성능들과 비교한 결과 제안된 형태 적응 필터가 천연 영상들과 resolution phantom 영상 모두에서 에지를 잘 보존하면서도 높은 신호대잡음비를 얻음을 확인하였다. In this paper, a local shape adaptive template filtering is proposed for an enhancement of signal-to-noise ratio in magnetic resonance imaging with little degradation of resolution. Uulike conventional filtering where template size and shape are fixed, multiple templates are defined and optimal template among the multiple templates is selected and applied in pixel-by-pixel basis. By the proposed scheme, edge blurring is minimized by selecting mostly matched template, yet SNR enhancement by the filtering is fully utilized. Compared to existing two-dimensional linear least square error Inter or direction-adaptive recursive filter, the proposed shape-adaptive filter with multiple templates provides higher SNR and sharper edges for both natural images and artificial resolution phantom images.

향상된 정수 기반 웨이블릿 변환을 이용한 무손실에 가까운 영상 압축

안창범(Ahn C . B .),송영철(Song Y . C .) 한국화상학회 1997 한국화상학회지 Vol.3 No.1

N/A An enhanced version of integer-based wavelet transform(TS Transform) coding is proposed for near lossless image compression. Due to flexibility in representing nonstationary image signals in time and frequency domains and ability in adapting to human visual characteristics, wavelet transform has unique advantages over other transforms currently used in image compression. In the Integer-based wavelet transform compression, original image is decomposed into multi-scalar bands. For the lowest band, a predictor is designed and error signal is entropy coded. For higher scalar bands, runlength coding for zero runs is employed with Huffman coding. In order to perform near lossless compression, pre-quantization of the original image is applied and the quantized image is losslessly encoded. To overcome visual degradation resulting from pre-quantization, a new pre-quantization scheme using multiple quantization tables is proposed. From simulation (512×512 size, 256 graylevel image) reconstructed images by the proposed algorithm show higher peak signal-to-noise ratio (PSNR) compared to those obtained by a float-point-based wavelet transform and the JPEG. Peak error in the reconstructed images by the integer-based wavelet transform is also much lower than those obtained by the JPEG and the float-point-based wavelet transform.

정수기반 웨이블릿 변환을 이용한 통합된 유 / 무손실 의학영상압축

안창범(Ahn C . B .),송영철(Song Y . C .),고형화(Koh H . H .),안용현(Ahn Y . H .) 한국화상학회 1998 한국화상학회지 Vol.4 No.1

완전 복원이 가능한 정수 기반 웨이블릿 변환을 사용한 통합된 유/무손실 압축을 제안하였다. 제안한 알고리즘은 선양자화 과정과 정수 기반 웨이블릿 변환, 그러고 QM 코더에 의한 웨이블릿 밴드들의 비트 평면 부호화에 기반을 두고 있다. 본 논문에서는 선양자화과정을 거친 후 그레이레벨의 수가 원영상과 일치하는 향상된 선양자화과정과, 웨이블릿 대역을 주파수 및 bit 평면에서 최적의 순서로 부호화하는 방법을 제안하였다. 모의 실험에서 , 제안된 알고리즘에 의해 재구성된 명상들은 부동 소수점 기반 웨이블릿 변환(Daubechies)과 JPEG에 의해 재구성된 영상들에 비해 높은 신호대잡음비를 보였다. 또한 정수 기반 웨이블릿 변환에 의하여 재구성된 영상들에서 최대에러는 JPEG과 부동 소수점 기반 웨이블릿 부호화방식에 비하여 훨씬 낮게 나타났다. 무손실에 가까운 압축에서, 선양자화된 영상에서의 웨이블릿 계수들의 무손실 부호화는 선양자화 없이 원영상의 웨이블릿 계수들의 비트 평면-밴드 정렬에 의해 비트율을 조절하는 부호화 방법에 비해 높은 PSNR을 보였다. 선양자화과정을 사용함으로써 최대에러를 조절할 수 있었으며, 이것은 의학영상압축에서 유용한 특징이 될 것으로 기대된다. A unified lossy to lossless compression algorithm is developed using integer-based reversible wavelet transforms. The proposed algorithm is based on a pre-quantization, integer-based wavelet transform, and a bit plane coding of the wavelet bands with QM coder. An improved pre-quantization scheme is proposed with which number of gray levels after pre-quantization is the same as that in the original image. An optimal encoding order for the wavelet bands in both spatial frequency domain and bit planes are devised. Prom simulation, reconstructed image by the proposed algorithm shows higher peak signal-to-noise ratio (PSNR) compared to those obtained by a float-point-based wavelet transform (Daubechies) and the JPEG. Peak errors in the reconstructed images by the integer-based wavelet transform are also much lower than those obtained by the JPBG and the float-point-based wavelet coding. In near lossless compression, lossless encoding of the wavelet coefficients of the pre-quantized image shows higher PSNR than those with bit rate control by the bit plane-band ordering of the wavelet coefficients without pre-quantization. Using pre-quantization, peak error can be easily controlled, which may be an important feature in medical image compression.

개량차분법에 의한 임의의 형상의 열처리제품의 급냉경화 해석

김계수,송영범,홍준표 ( K . S . Kim,Y . B . Song,C . P . Hong ) 한국열처리공학회 1989 熱處理工學會誌 Vol.2 No.2

Understanding and control of thermal history of heat-treated components are very important in heat treating process. A computer program was developed for prediction of thermal history in quenching process with various cooling media and agitation conditions. Computer simulation of heat flow in quenching treatment of spur gear of SCM 22 H was carried out by two and three dimensional finite difference method. Distributions of microstructure and hardness in heat-treated spur gear were predicted by computer simulation, and the results showed a good agreement with the experiments. It was concluded that the

7175 합금 단조재의 미세조직과 기계적 성질에 미치는 중간가공열처리 영향

이용연,송영범,손영일,이경훈,은일상 ( Y . Y . Lee,Y . B . Song,S . T . Im,Y . I . Son,K . H . Lee,I . S . Eun ) 한국열처리공학회 1997 熱處理工學會誌 Vol.10 No.3

The microstructure, tensile and impact properties of forgings of 7175 aluminium alloy have been studied as a function of intermediate thermo-mechanical treatment(ITMT) process. The ITMT process is consisted of warm working and recrystallization. In the case that the billet was warm-worked above 60% below 250℃ and recrystallized at 475℃, the grain size revealed about 17㎛ which corresponds to one third of that of conventional process. The refinement of grain size leaded to the improved ductility and impact energy without sacrifice of tensile strength. It was found that the ITMT processed specimen behaved isotropically due to the near equiaxed grains. It was observed that the ITMT processed specimen showed the mixed fracture mode of transgranular and intergranular, instead of intergranular mode. This change of fracture mode contributed to the improved ductility and impact property.

극초고강도 이차경화형 마르텐사이트강의 기계적성질에 미치는 오스포밍 공정의 영향

김수빈 ( S. B. Kim ),원윤정 ( Y. J. Won ),송영범 ( Y. B. Song ),조기섭 ( K. S. Cho ) 한국열처리공학회 2021 熱處理工學會誌 Vol.34 No.4

Two types of secondary hardening martensitic steels, 10Co-14Ni and 6Co-5Ni, were produced by vacuum induction melting to investigate the effect of ausforming process on mechanical properties. According to the results of present study, the alloy samples ausformed at low temperature indicated a rather low hardness level in overall aging time despite the refinement of martensite lath width. As the result can closely be related with the presence of primary carbides precipitated within the initial austenite matrix, we confirmed that, in ultrahigh strength secondary hardening martensitic alloy steels, the ausforming process can rather limit the degree of secondary hardening during the subsequent aging treatment. (Received July 8, 2021; Revised July 15, 2021; Accepted July 21, 2021)

인공지능을 활용한 기계학습 앙상블 모델 개발

이근원,원윤정,송영범,조기섭,Lee, K.W.,Won, Y.J.,Song, Y.B.,Cho, K.S. 한국열처리공학회 2021 熱處理工學會誌 Vol.34 No.5

To predict mechanical properties of secondary hardening martensitic steels, a machine learning ensemble model was established. Based on ANN(Artificial Neural Network) architecture, some kinds of methods was considered to optimize the model. In particular, interaction features, which can reflect interactions between chemical compositions and processing conditions of real alloy system, was considered by means of feature engineering, and then K-Fold cross validation coupled with bagging ensemble were investigated to reduce R2_score and a factor indicating average learning errors owing to biased experimental database.