• 정보처리학회논문지B
• /
• 제15B권5호
• /
• pp.435-448
• /
• 2008

사용자 포즈의 3차원 데이터 생성을 통한 3차원 포즈 인식은 2차원 포즈 인식의 문제점을 해결하기 위해서 많이 연구되고 있지만, 3차원 표면 데이터의 방대한 양으로 포즈 인식에서 중요한 특징 추출(feature extraction)이 어렵고 수행 시간이 많이 걸리는 문제점을 가지고 있다. 본 논문에서는 3차원 포즈 인식의 두 가지 문제점인 특징 추출의 어려움과 느린 처리속도를 개선하기 위해서 3차원 형상복원 기술로 모델의 3차원 표면 점들로 구성된 데이터를 2차원 데이터로 변환하는 차원 축소(dimension reduction) 방법을 제안한다. 실린더형 외곽점을 이용한 메쉬없는 매개변수화(meshless parameterization) 방법은 방대한 데이터인 3차원 포즈 데이터를 2차원 데이터로 변환하여 특징 추출과 매칭과정의 연산 속도를 향상 시키며, 특징 추출의 효율성 검증을 위해 간단한 환경에서 실험이 가능한 손 포즈 인식 및 인간 포즈 인식에 적용하였다.

• 전자공학회논문지CI
• /
• 제49권4호
• /
• pp.42-53
• /
• 2012

패턴 인식과 기계 학습 분야에서 분류는 가장 기본적으로 해결해야 하는 문제의 유형이다. Adaboost 알고리즘은 Boosting 알고리즘의 아이디어를 실제 데이터분석에 이용할 수 있도록 개량한 방법으로써, 단계를 반복하여 나온 여러 개의 약한 분류기와 가중치 값들의 조합으로 강한 분류기를 생성하는 두 개의 클래스를 분류하는 분류기이다. 주성분 분석법과 선형 판별 분석법은 높은 차원의 특징 벡터를 낮은 차원의 특징 벡터로 축소하는 특징 벡터의 차원 감소와 데이터의 특징 추출에도 유용하게 사용되는 방법들이다. 본 논문에서는, 주성분 분석법과 선형 판별 분석법을 이용하여 추출한 특징을 Adaboost 알고리즘의 약 분류기로 사용함으로써, 특징 추출과 분류를 동시에 하고, 인식률을 높이는 효율적인 Boosted-PCA와 Boosted-LDA 알고리즘을 제안한다. 마지막 장에서는, 제안하는 알고리즘으로 UCI Data-Set 중 2 Class-Data와 FRGC Data의 남자와 여자 영상에 대해서 분류 실험을 진행하였다. 실험의 결과로 제안한 Boosted-PCA와 Boosted-LDA 알고리즘이 기존의 특징 추출 알고리즘과 최근접 이웃 분류기, SVM을 이용한 분류기 방법과 비교하여 인식률이 향상됨을 보인다.

• 한국컴퓨터정보학회논문지
• /
• 제21권4호
• /
• pp.55-62
• /
• 2016

In this paper, we propose the INDEM(Internet News Data Extraction Middleware) system for the removal of the unnecessary data in internet news. Although data on the internet can be used in various fields such as source of data of IR(Information Retrieval), Data mining and knowledge information service, it contains a lot of unnecessary information. The removal of the unnecessary data is a problem to be solved prior to the study of the knowledge-based information service that is based on the data of the web page. The INDEM system parses html and explores the XPath, and it is to perform the analysis. The user simply utilize INDEM by implementing an abstract class that provides INDEM, and can obtain the analysis information. INDEM System through this process delivers the analysis information including the main contents of news site to the users. In this paper, the INDEM system was adapted in a stand-alone and web service system and it was evaluated on the basis of 16 news site. As a result, performance of the INDEM system is affected in html source data size and complexity of used html grammar than the main news data size.

• Journal of Ginseng Research
• /
• 제40권3호
• /
• pp.229-236
• /
• 2016

Background: Extraction conditions greatly affect composition, as well as biological activity. Therefore, optimization is essential for maximum efficacy. Methods: Korean Red Ginseng (KRG) was extracted under different conditions and antioxidant activity, extraction yield, and ginsenoside Rg1 and phenolic content evaluated. Optimized extraction conditions were suggested using response surface methodology for maximum antioxidant activity and extraction yield. Results: Analysis of KRG extraction conditions using response surface methodology showed a good fit of experimental data as demonstrated by regression analysis. Among extraction factors, such as extraction solvent and extraction time and temperature, ethanol concentration greatly affected antioxidant activity, extraction yield, and ginsenoside Rg1 and phenolic content. The optimal conditions for maximum antioxidant activity and extraction yield were an ethanol concentration of 48.8%, an extraction time 73.3 min, and an extraction temperature of $90^{\circ}C$. The antioxidant activity and extraction yield under optimal conditions were 43.7% and 23.2% of dried KRG, respectively. Conclusion: Ethanol concentration is an important extraction factor for KRG antioxidant activity and extraction yield. Optimized extraction conditions provide useful economic advantages in KRG development for functional products.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권12호
• /
• pp.6038-6053
• /
• 2017

In this study, we propose an automatic melody extraction algorithm using deep learning. In this algorithm, feature images, generated using the energy of frequency band, are extracted from polyphonic audio files and a deep learning technique, a convolutional neural network (CNN), is applied on the feature images. In the training data, a short frame of polyphonic music is labeled as a musical note and a classifier based on CNN is learned in order to determine a pitch value of a short frame of audio signal. We want to build a novel structure of melody extraction, thus the proposed algorithm has a simple structure and instead of using various signal processing techniques for melody extraction, we use only a CNN to find a melody from a polyphonic audio. Despite of simple structure, the promising results are obtained in the experiments. Compared with state-of-the-art algorithms, the proposed algorithm did not give the best result, but comparable results were obtained and we believe they could be improved with the appropriate training data. In this paper, melody extraction and the proposed algorithm are introduced first, and the proposed algorithm is then further explained in detail. Finally, we present our experiment and the comparison of results follows.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제3권2호
• /
• pp.194-199
• /
• 2003

Data are an expression of the language or numerical values that show some features. And the information is extracted from data for the specific purposes. The knowledge is utilized as information to construct rules that recognize patterns or make a decision. Today, knowledge extraction and application of that are broadly accomplished for the easy comprehension and the performance improvement of systems in the several industrial fields. The knowledge extraction can be achieved by some steps that include the knowledge acquisition, expression, and implementation. Such extracted knowledge is drawn by rules with data mining techniques. Clustering (CL), input space partition (ISP), neuro-fuzzy (NF), neural network (NN), extension matrix (EM), etc. are employed for the knowledge expression based upon rules. In this paper, the various approaches of the knowledge extraction are surveyed and categorized by methodologies and applied industrial fields. Also, the trend and examples of each approaches are shown in the tables and graphes using the categories such as CL, ISP, NF, NN, EM, and so on.

• 대한원격탐사학회지
• /
• 제8권2호
• /
• pp.105-123
• /
• 1992

A study was made on the feature extraction for non-metallic one deposits and their related geology using the Remote Sensing and Airborne Radiometric data. The area chosen is around the Haenam area, where dickite and Quarzite mines are distributed in. The geology of the area consists mainly of Cretaceous volcanics and PreCambrian metamorphic. The methods applied are study on the reflectance characteristics of minerals and rocks sampled in the study area, and the feature extraction extraction of histogram normalized images for Landsat TM and Airborne Radiometric data, and finally evaluation of applicability of some useful pattern recognition techniques for regional lithological mapping. As a result, reflectances of non-metallic minerals are much higher than rock samples in the area. However, low grade dickites are slightly higher than rock samples, probably due to their greyish colour and also their textural features which may scatter the reflectance and may be capable of capturing much hychoryl ions. The reflectances of rock samples may depend on the degree of whiteness of samples. The outcrops or mine dumps in the study area were most effectively extracted on the histogram normalized image of TM Band 1, 2 and 3, due to their high reflectivity. The Masking technique using the above bands may be the most effective and the natural colour composite may provide some success as well. The colour composite image of PCA may also be effective in extracting geological features, and airborne radiometric data may be useful to some degree as an complementary tool.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2011년도 추계학술발표대회
• /
• pp.355-356
• /
• 2011

LiDAR data is very large, which contains an amount of redundant information. The information not only takes up a lot of storage space but also brings much inconvenience to the LIDAR data transmission and application. Therefore, a simplified method was proposed for LiDAR data based on quad-tree structure in this paper. The boundary contour lines of the buildings are displayed as building extraction. Experimental results show that the method is efficient for point's simplification according to the rule of mapping.

• 한국측량학회지
• /
• 제27권1호
• /
• pp.723-733
• /
• 2009

This study proposes a sequence of procedures to extract building boundaries and planar patches through segmentation of rasterized lidar data. Although previous approaches to building extraction have been shown satisfactory, there still exist needs to increase the degree of automation. The methodologies proposed in this study are as follows: Firstly, lidar data are rasterized into grid form in order to exploit its rapid access to neighboring elevations and image operations. Secondly, propagation of errors in raw data is taken into account for in assessing the quality of gradients-driven properties and further in choosing suitable parameters. Thirdly, extraction of planar patches is conducted through a sequence of processes: histogram analysis, least squares fitting, and region merging. Experimental results show that the geometric components of building models could be extracted by the proposed approach in a streamlined way.

• 한국지능시스템학회논문지
• /
• 제8권3호
• /
• pp.16-25
• /
• 1998

본 논문에서는 초 다중 밴드 환경의 효과적인 데이터 분류를 위해서 Roungh 집합 이론을 이용한 특징 추출 방법을 제안한다. 다중 분광 이미지 데이터의 특성을 분석하고, 그 분석 결과를 토대로 Rough집합이론의 식별 능력을 이용하여 가장 효과적인 밴드를 선택할 수 있도록 한다. 실험으로는 Landsat TM으로부터 취득한 데이터에 적용시켰으며, 이를 통해 전통적인 밴드 특성에 의한 밴드 선택 방법과 본 논문에서 제안하는 러프 집합 이론을 이용한 밴드 선택 방법이 일치됨을 보이고 이를 통해 초다중 밴드 환경에서의 특징 추출에 대한 이론적 근거를 제시한다.