• 제목/요약/키워드: Random extraction

검색결과 202건 처리시간 0.028초

Active Learning과 군집화를 이용한 고정키어구 추출 (Keyphrase Extraction Using Active Learning and Clustering)

  • 이현우;차정원
    • 대한음성학회지:말소리
    • /
    • 제66호
    • /
    • pp.87-103
    • /
    • 2008
  • We describe a new active learning method in conditional random fields (CRFs) framework for keyphrase extraction. To save elaboration in annotation, we use diversity and representative measure. We select high diversity training candidates by sentence confidence value. We also select high representative candidates by clustering the part-of-speech patterns of contexts. In the experiments using dialog corpus, our method achieves 86.80% and saves 88% training corpus compared with those of supervised method. From the results of experiment, we can see that the proposed method shows improved performance over the previous methods. Additionally, the proposed method can be applied to other applications easily since its implementation is independent on applications.

  • PDF

랜덤한 점분포를 가진 영상을 사용한 워터마킹의 견고성 (A Robustness of Watermarking Using Random Dot Distribute Image)

  • 이인정
    • 한국정보기술응용학회:학술대회논문집
    • /
    • 한국정보기술응용학회 2002년도 추계공동학술대회 정보환경 변화에 따른 신정보기술 패러다임
    • /
    • pp.66-71
    • /
    • 2002
  • 삽입하려는 암호영상이 원본 전체 이미지 내에 랜덤하게 분포할수록 삽입과 추출성능이 좋아지는데 본 논문에서는 랜덤성이 우수한 오토스테레오그램을 삽입암호 영상으로 사용하여 워터마크 하였을 때 복원과 추출성능이 양호함을 알아보았고 이를 이용하여 로고영상을 워터마크하려는 영상의 전 영역에 랜덤하게 고루 분포하게하여 워터마크한 후 역으로 변환하여 로고를 찾았다. 영상이 많이 손상되었을 때도 추출된 로고가 육안으로 식별할 수 있을 정도로 견고성이 우수함을 알아보았다.

  • PDF

랜덤한 점분포를 가진 영상을 사용한 워터마킹의 견고성 (A Robustness of Watermarking Using Random Dot Distribute Image)

  • 이인정
    • 한국산업정보학회:학술대회논문집
    • /
    • 한국산업정보학회 2002년도 추계공동학술대회
    • /
    • pp.66-71
    • /
    • 2002
  • 삽입하려는 암호영상이 원본 전체 이미지 내에 랜덤하게 분포할수록 삽입과 추출성능이 좋아지는데 본 논문에서는 랜덤성이 우수한 오토스테레오그램을 삽입암호 영상으로 사용하여 워터마크 하였을 때 복원과 추출성능이 양호함을 알아보았고 이를 이용하여 로고영상을 워터마크 하려는 영상의 전 영역에 랜덤하게 고루 분포하게 하여 워터마크 한 후 역으로 변환하여 로고를 찾았다. 영상이 많이 손상되었을 때도 추출된 로고가 육안으로 식별할 수 있을 정도로 견고성이 우수함을 알아보았다.

  • PDF

Coastline Extraction Using RTK GPS/GLONASS Combination Method

  • 서동주
    • 한국해양공학회지
    • /
    • 제20권3호
    • /
    • pp.15-23
    • /
    • 2006
  • On this study, realtime GPS technique and combination of GPS/GLONASS technique are used to extracting coastline. Th£ object of coastline is Gwanganri beach located in Busan. The coastline is observed along the traces of coastline when high wave of scar by using digital map of 1:1,000 and at random time zone, coastline is surveyed along the boundary line that is contacting with sea water level. When the coastline of random time zone is converted by height of tide table, the coastline when high wave of scar and converted coastline are shown as coincident approximately.

Quadtree 구조 및 프랙탈 특성을 이용한 Hyperion 영상의 노이즈 밴드 추출 (Noise Band Extraction of Hyperion Image using Quadtree Structure and Fractal Characteristic)

  • 장안진;김용일
    • 대한원격탐사학회지
    • /
    • 제26권5호
    • /
    • pp.489-495
    • /
    • 2010
  • 초분광 영상은 넓은 범위의 파장 영역의 유용한 정보를 많은 수의 밴드를 통해 취득한다. 하지만, 인접 밴드 간의 상관관계, 계산량, 노이즈로 인해 전처리없이 활용할 경우 부정확한 결과를 도출한다. 따라서 영상에서 노이즈 밴드 추출하여 제거하는 작업이 반드시 필요하다. 기존의 연구들은 영상 전체에 대한 특성치 만을 이용하였기 때문에 영상의 국지적 특성을 고려해야 한다. 본 연구에서는 Hyperion 영상을 대상으로 하였으며, 자료구조 기법 중 하나인 Quadtree와 이용하여 노이즈 밴드를 추출하였다. Quadtree 구조로 분할된 영역의 프랙탈 차원을 계산하고 프랙탈 차원의 분산을 이용하였다. Hyperion 영상에 존재하는 노이즈 종류 중 무작위 노이즈를 포함하고 있는 밴드 추출에 초점을 맞추었으며, 시각적으로 판단하여 작성한 참조자료와 비교하였다. 제안된 알고리즘 적용 결과 무작위 노이즈가 포함된 밴드 대부분이 추출되었으며, 영상에 관계없이 30개 이상의 노이즈 밴드를 제거할 수 있음을 확인하였다.

Background Subtraction using Random Walks with Restart

  • Kim, Tae-Hoon;Lee, Kyoung-Mu;Lee, Sang-Uk
    • 한국방송∙미디어공학회:학술대회논문집
    • /
    • 한국방송공학회 2009년도 IWAIT
    • /
    • pp.63-66
    • /
    • 2009
  • Automatic segmentation of foreground from background in video sequences has attracted lots of attention in computer vision. This paper proposes a novel framework for the background subtraction that the foreground is segmented from the background by directly subtracting a background image from each frame. Most previous works focus on the extraction of more reliable seeds with threshold, because the errors are occurred by noise, weak color difference and so on. Our method has good segmentations from the approximate seeds by using the Random Walks with Restart (RWR). Experimental results with live videos demonstrate the relevance and accuracy of our algorithm.

  • PDF

Random Amplified Polymorphic DNA (RAPD) Identification of Genetic Variation in Chlorella species

  • CHO Jung Jong;KIM Yong-Tae;HUR Sung Bum;KIM Young Tae
    • 한국수산과학회지
    • /
    • 제29권6호
    • /
    • pp.761-769
    • /
    • 1996
  • The random amplified polymorphic DNA (RAPD) technique was used to characterize 18 reference strains of microalgae, mostly Chlorella species, collected from various localities around Korea peninsular. Eighteen strains consist of four genera of the family marine Chlorella from 12 samples, two genera of fresh water Chlorella from three samples, and three genera on Nannochloris. Twenty 10-mer anonymous primers were screened for amplification of genomic DNA extracted from samples using the CTAB extraction method. Nineteen of these oligonucleotide primers were positive or band producing. Three of 20 random primers (OPA 10, OPA 12, and OPA 18) resulted in both clear band and a high degree of reproducibility and showed some potential to be used to discriminate individual samples of both genetically hetero-and homogeneous populations, in determining phylogenetic relationships between species within a genus and developing individual fingerprints for each samples.

  • PDF

A New Approach to Fingerprint Detection Using a Combination of Minutiae Points and Invariant Moments Parameters

  • Basak, Sarnali;Islam, Md. Imdadul;Amin, M.R.
    • Journal of Information Processing Systems
    • /
    • 제8권3호
    • /
    • pp.421-436
    • /
    • 2012
  • Different types of fingerprint detection algorithms that are based on extraction of minutiae points are prevalent in recent literature. In this paper, we propose a new algorithm to locate the virtual core point/centroid of an image. The Euclidean distance between the virtual core point and the minutiae points is taken as a random variable. The mean, variance, skewness, and kurtosis of the random variable are taken as the statistical parameters of the image to observe the similarities or dissimilarities among fingerprints from the same or different persons. Finally, we verified our observations with a moment parameter-based analysis of some previous works.

Feature Extraction of Non-proliferative Diabetic Retinopathy Using Faster R-CNN and Automatic Severity Classification System Using Random Forest Method

  • Jung, Younghoon;Kim, Daewon
    • Journal of Information Processing Systems
    • /
    • 제18권5호
    • /
    • pp.599-613
    • /
    • 2022
  • Non-proliferative diabetic retinopathy is a representative complication of diabetic patients and is known to be a major cause of impaired vision and blindness. There has been ongoing research on automatic detection of diabetic retinopathy, however, there is also a growing need for research on an automatic severity classification system. This study proposes an automatic detection system for pathological symptoms of diabetic retinopathy such as microaneurysms, retinal hemorrhage, and hard exudate by applying the Faster R-CNN technique. An automatic severity classification system was devised by training and testing a Random Forest classifier based on the data obtained through preprocessing of detected features. An experiment of classifying 228 test fundus images with the proposed classification system showed 97.8% accuracy.

Effect of extraction conditions on the stability and safety of sericin

  • Ji Hae, Lee;Hyun-Bok, Kim;HaeYong, Kweon
    • International Journal of Industrial Entomology and Biomaterials
    • /
    • 제45권2호
    • /
    • pp.93-98
    • /
    • 2022
  • To assess the feasibility of silk sericin for non-textile application, the storage stability and biological safety of sericin were examined. It was extracted at 37℃, 70℃, 100℃, and 121℃ for 1, 3, and 5 h to elucidate the effect of extraction condition on the stability and safety of silk sericin. The solubility was increased till approximately 26% with extraction temperature of 121℃ for 1 h. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) showed that the molecular weight distribution depended on the extraction conditions. Extracted sericin displayed typical UV absorption bands upon spectrometric analysis. To examine the reproducibility of its obtained conformation, sericin was extracted thrice and its circular dichroism (CD) spectra was measured each time. Most CD spectra showed reproducibility regardless of temperature and time except under 100℃ extraction condition. The diversity of CD spectrum showed gradual reduction and was finally coincident with extraction time from 1 to 5 h. Notably, sericin has a negative peak of approximately 200 nm attributed to random coil conformation, regardless of extraction condition. However, at the 100℃ extraction condition, sericin showed both bands to be negative bands of approximately 200 and 220 nm, respectively. Sericin was centrifuged to determine the stability of storage conditions. The sericin extracted at 100℃ and 121℃ for 1 h was found to form gel rapidly within 1 h, but at 121℃ condition, the gel fraction was approximately 20% within 1 h which retained its phase regardless of storage time. The gel fraction of sericin extracted at 100℃ for 5 h increased with time, however at the 121℃ for 5 h condition, the gel fraction was measured to be less than 10% regardless of increase in storage time. PetriflimTM AC plates test showed that sericin was safe from aerobic bacteria activity by extraction under high temperature.