• 한국정보통신학회논문지
• /
• 제24권11호
• /
• pp.1424-1430
• /
• 2020

본 논문에서는 영상에 존재하는 잡음의 특성을 고려하여 AWGN을 제거하기 위한 알고리즘을 제안한다. 제안한 알고리즘은 출력 계산을 위해 블록 매칭을 사용하였으며, 센터 마스크와 매칭 마스크의 유사도 판별하여 추정치를 계산한다. 필터의 출력은 추정치와 입력 화소값을 가감하여 계산하며, 센터 마스크의 표준 편차와 잡음 상수에 따라 가중치를 부여하여 최종 출력을 구한다. 제안하는 알고리즘을 평가하기 위해 기존 방법들과 비교하여 시뮬레이션하였으며, 확대영상 및 PSNR비교를 통해 분석하였다. 제안한 알고리즘은 잡음의 영향을 최소화하였으며, 영상의 중요 특성을 보존하며 효율적으로 잡음을 제거하는 성능을 보였다.

• 한국소프트웨어감정평가학회 논문지
• /
• 제17권1호
• /
• pp.25-30
• /
• 2021

임베디드 시스템 소스코드 감정목적물의 유사성을 판단하는 경우, 제공된 소스코드가 컴파일이 가능한 상태인지, 실행이 정상적으로 되는 것인지, 하드웨어와의 인터페이스가 일치하는지 등의 확인이 되지 않은 상태로 제공되는 경우가 지속적으로 발생하고 있다. 분쟁을 제기하는 측에서는 동작의 특성이 유사하고, 나타나는 기능의 효과가 유사한 이유로 소스코드의 많은 부분이 유사할 것으로 판단하고, 감정을 요청하게 되지만, 위의 여러 가지 상황으로 유사성 분석 결과가 기대와 다르게 나타나는 가능성이 우려된다. 본 연구에서는 감정사례를 통해 소스코드의 분석 과정과 검증되지 않은 소스코드의 유사성 도출의 개선방향을 제시한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권8호
• /
• pp.2895-2912
• /
• 2014

Content-based image retrieval has been the most important technique for managing huge amount of images. The fundamental yet highly challenging problem in this field is how to measure the content-level similarity based on the low-level image features. The primary difficulties lie in the great variance within images, e.g. background, illumination, viewpoint and pose. Intuitively, an ideal similarity measure should be able to adapt the data distribution, discover and highlight the content-level information, and be robust to those variances. Motivated by these observations, we in this paper propose a probabilistic similarity learning approach. We first model the distribution of low-level image features and derive the free energy kernel (FEK), i.e., similarity measure, based on the distribution. Then, we propose a learning approach for the derived kernel, under the criterion that the kernel outputs high similarity for those images sharing the same class labels and output low similarity for those without the same label. The advantages of the proposed approach, in comparison with previous approaches, are threefold. (1) With the ability inherited from probabilistic models, the similarity measure can well adapt to data distribution. (2) Benefitting from the content-level hidden variables within the probabilistic models, the similarity measure is able to capture content-level cues. (3) It fully exploits class label in the supervised learning procedure. The proposed approach is extensively evaluated on two well-known databases. It achieves highly competitive performance on most experiments, which validates its advantages.

• Asian-Australasian Journal of Animal Sciences
• /
• 제18권4호
• /
• pp.483-489
• /
• 2005

To understand the molecular mechanisms that regulate intramuscular fat deposition and its release, cDNA clones expressed in adipose tissues of Korean cattle were identified by differential screening from adipose tissue cDNA library. By partial nucleotide sequencing of 486 clones and a search for sequence similarity in NCBI nucleotide databases, 245 clones revealed unique clones. By a functional grouping of the clones, 14% of the clones were categorized to metabolism and enzyme-related group (stearoyl CoA desaturase, lactate dehydrogenase, fatty acid synthase, ATP citrate lyase, lipoprotein lipase, acetyl CoA synthetase, etc), and 6% to signal transduction/cell cycle-related group (C/EBP, cAMP-regulated phosphoprotein, calmodulin, cyclin G1, cyclin H, etc), and 4% to cytoskeleton and extracellular matrix components (vimentin, ankyrin 2, gelosin, syntenin, talin, prefoldin 5). The obtained 245 clones will be useful to study lipid metabolism and signal transduction pathway in adipose tissues and to study obesity in human. Some clones were subjected to full-sequencing containing open reading frame. The cDNA clone of bovine homolog of human prefoldin 5 gene had a total length of 959 nucleotides coding for 139 amino acids. Comparison of the deduced amino acid sequences of bovine prefoldin 5 with those of human and mouse showed over 95% identity. The cDNA clone of bovine homolog of human ubiquitin-like/S30 ribosomal fusion protein gene had a total length of 484 nucleotides coding for 133 amino acids. Comparison of the deduced amino acid sequences of bovine ubiquitin-like/S30 ribosomal fusion protein gene with those of human, rat and mouse showed over 97% identity. The cDNA clone of bovine homolog of human proteolipid protein 2 mRNA had a total length of 928 nucleotides coding for 152 amino acids. Comparison of the deduced amino acid sequences of bovine proteolipid protein 2 with those of human and mouse showed 87.5% similarity. The cDNA clone of bovine homolog of rat thymosin beta 4 had a total length of 602 nucleotides coding for 44 amino acids. Comparison of the deduced amino acid sequences of bovine thymosin beta 4 gene with those of human, mouse and rat showed 93.1% similarity. The cDNA clone of bovine homolog of human myotrophin mRNA had a total length of 790 nucleotides coding for 118 amino acids. Comparison of the deduced amino acid sequences of bovine myotrophin gene with those of human, mouse and rat showed 83.9% similarity. The functional role of these clones in adipose tissues needs to be established.

• 한국산업융합학회 논문집
• /
• 제23권3호
• /
• pp.523-529
• /
• 2020

Foot drop is a common symptom in stroke patients due to central nervous system (CNS) damage, which causes walking disturbances. Functional electrical stimulation (FES) is an effective rehabilitation method for stroke patients with CNS damage. Aim of this study was to determine the effectiveness of 6 weeks FES walking training based lower limb muscle synergy of stroke patients. Lower limb muscle synergies were extracted from electromyography (EMG) using a non-negative matrix factorization algorithm (NMF) method. Cosine similarity and cross correlation were calculated for similarity comparison with healthy subjects. In both stroke patients, the similarity of leg muscle synergy during walking changed to similar to that of healthy subjects due to a decrease in foot drop during. FES walking intervention influenced the similarity of muscle synergies during walking of stroke patients. This intervention has an effective method on foot drop and improving the gait performance of stroke patients.

• 전기전자학회논문지
• /
• 제24권4호
• /
• pp.1109-1116
• /
• 2020

사람이 따라 하는 능력을 평가하는 스코어는 아동의 발달 단계 혹은 골프, 무용 동작 등을 점검하는 좋은 수단이 될 수 있다. 또한, 이는 AR, VR 응용에서 HCI로도 활용될 수 있다. 본 논문에서는 동작을 주도해서 수행하는 시범자와 그 동작을 따라 하는 참여자 간의 동작 유사도를 평가하는 방안을 제시하고, 여기서 우리는 Openpose의 키포인트 벡터 유사도의 유클리디안 L2 거리를 활용하는 동작 유사도를 제안한다. 제안된 기법은 DTW를 사용하기 때문에 시간 지연차가 있는 동작에 유연하게 대처할 수 있다.

• 한국멀티미디어학회논문지
• /
• 제25권8호
• /
• pp.1059-1068
• /
• 2022

Recently, research using image-based deep learning is being conducted to determine similar fonts or recommend fonts. In order to increase the accuracy in judging the similarity of Hangul fonts, a previous study was conducted to calculate the similarity according to the combination of stroke elements. In this study, we tried to solve this problem by designing an integrated model that reflects the weights for each stroke element. By comparing the results of the user's font similarity calculation conducted in the previous study and the weighted model, it was confirmed that there was no difference in the ranking of the influence of the stroke elements. However, as a result of comparison by letter sizes, it was confirmed that there was a difference in the ranking of the influence of stroke elements. Accordingly, we proposed a weighted model set separately for each font size.

• 대한수학회보
• /
• 제41권3호
• /
• pp.435-450
• /
• 2004

We show that two concepts of h-stability and h-stability in variation for nonlinear difference systems are equivalent by using the concept of $n_{\infty}$-summable similarity of their associated variational systems. Also, we study h-stability for perturbed non-linear system y(n+1) =f(n,y(n)) + g(n,y(n), Sy(n)) of nonlinear difference system x(n+1) =f(n,x(n)) using the comparison principle and extended discrete Bihari-type inequality.

• 한국정보기술학회 영문논문지
• /
• 제9권2호
• /
• pp.65-74
• /
• 2019

In this paper, we describe a new swing analysis system. We design this system to provide various information about golf swings and to help to correct wrong swings. We visualize three-dimensional skeletal information obtained from Kinect through various views. Golfers can see their swing behavior through these views. This system can calculate the similarity between the two trajectories obtained from Kinect to determine the similarity of swing trajectories of different golfers. Input trajectories are resampled to have equal spacing and are performed scaling and translation for accurate trajectory comparison. We have verified the usefulness of the proposed system through various analyzes.

• 한국균학회지
• /
• 제25권1호통권80호
• /
• pp.20-25
• /
• 1997

느타리버섯 균주(Pleurotus isolates)와 이들 균주들을 교잡하여 얻어진 균사에서 추출한 단백질을 전기영동하여 균주간의 유전적 유연관계를 밝힌 결과, P. ostreatus 201은 자가교잡 균주 혹은 타가교잡 균주들과 낮은 유연관계를 나타냈다. P. ostreatus 2042도 역시 동일한 결과를 나타냈다. 그러나, P. ostreatus $2042{\times}P$. ostreatus 202, P. ostreatus $2042{\times}P$. sajor-caju, 그리고 ostreatus $2042{\times}P$ ostreatus 900은 높은 상관관계를 나타냈다. P. ostreatus 202와 P. ostreatus 900은 자가교잡 균주 혹은 타가 교잡된 균주들과 낮은 상관관계를 나타내었다. 그러나 자가 교잡된 P. ostreatus와 P. $ostreatus{\times}P$. florida, 그리고 P. florida와 자가교잡된 P. florida는 각각 서로 높은 상관관계를 나타냈다.