• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 춘계학술발표회논문집
• /
• pp.267-273
• /
• 2004

The advantage of the remote sensing is extraction the information of wide area rapidly. Such advantage is the resource and environment are quick and efficient method to grasps accurately method through the land cover classification of wide area. Accordingly this study is used to the high-resolution (6.6m) Electro-Optical Camera (EOC) panchromatic image of the first Korea Multi-Purpose Satellite 1 (KOMPSAT-1) and the multi-spectral Moderate Resolution Imaging Spectroradiometer (MODIS) image data(36 bands).We accomplished FCM classification technique with MLC technique to be general land cover classification method in the content of research. And evaluated the accuracy assessment of two classification method.

• 대한원격탐사학회지
• /
• 제28권3호
• /
• pp.329-336
• /
• 2012

This paper give further assessment on the original DoP-CPD classification scheme. This paper provides some additional comparative study on the DoP-CPD with H/A/alpha classifier in terms of multi look effects and classification performances. The statistics and multi looking effects of the DoP and CPD were analyzed with measured polarimetric SAR data. DoP-CPD is less sensitive to the number of averaging pixels than the entropy-alpha technique. A DoP-CPD diagram with appropriate boundaries between six different classes was then developed based on the data analysis. A polarimetric SAR image DoP-CPD classification technique is verified with C-band polarimetric RADARSAT-2 images.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.467-469
• /
• 2003

This research aimed to develop forest type classification technique for the mixed forest with coniferous and broad-leaved species using the high resolution satellite data. QuickBird data was used as satellite data. The method of this research was to extract satellite data for every single tree crown using image segmentation technique, then to evaluate the accuracy of classification by changing grouping criteria such as tree species, families, coniferous or broad-leaved species, and timber prices. As a result, the classification of tree species and families level was inaccurate, on the other hand, coniferous or broad-leaved species and timber price level was high accurate.

• Journal of Computing Science and Engineering
• /
• 제5권2호
• /
• pp.150-160
• /
• 2011

Automatic text classification has a long history and many studies have been conducted in this field. In particular, many machine learning algorithms and information retrieval techniques have been applied to text classification tasks. Even though much technical progress has been made in text classification, there is still room for improvement in text classification. In this paper, we will discuss remaining issues in improving text classification. In this paper, three improvement issues are presented including automatic training data generation, noisy data treatment and term weighting and indexing, and four actual studies and their empirical results for those issues are introduced. First, the semi-supervised learning technique is applied to text classification to efficiently create training data. For effective noisy data treatment, a noisy data reduction method and a robust text classifier from noisy data are developed as a solution. Finally, the term weighting and indexing technique is revised by reflecting the importance of sentences into term weight calculation using summarization techniques.

• International Journal of Computer Science & Network Security
• /
• 제23권1호
• /
• pp.33-40
• /
• 2023

Machine-learning techniques are discovering effective performance on data analytics. Classification and regression are supported for prediction on different kinds of data. There are various breeds of classification techniques are using based on nature of data. Threshold determination is essential to making better model for unlabelled data. In this paper, threshold value applied as range, based on min-max normalization technique for creating labels and multiclass classification performed on rainfall data. Binary classification is applied on autism data and classification techniques applied on child abuse data. Performance of each technique analysed with the evaluation metrics.

• 한국측량학회지
• /
• 제22권2호
• /
• pp.171-178
• /
• 2004

원격탐사의 장점 중 하나는 넓은 지역의 정량적이고 정성적인 정보를 신속하게 추출할 수 있는 것이다. 그것은 넓은 지역의 토지피복을 분류하여 자원 및 환경을 신속하고 정확하게 파악하는 효과적인 수단이다. 따라서 본 연구에서는 알고리즘 개발을 통하여 더 나은 토지피복분류 방법을 제시하고자 하였다. 연구내용으로는 정형화된 토지피복분류방법인 최대우도법을 수행하고, 새로운 FCM 알고리즘을 이용한 영상분류를 수행하여 두 방법의 분류정확도를 비교 평가하였다. 또한 이용된 영상들은 한국항공우주연구원에서 매일 실시간으로 수신하고 있기 때문에 시간과 비용면에서 경제적인 위성영상을 이용하였다. 해상력은 다소 떨어지는 다파장대(36개 bands)의 MODIS 위성영상과 단 밴드인 KOMPSAT-1 EOC 위성영상을 이용하여 중합영상을 생성하여 토지피복분류에 이용하였다.

• 방송공학회논문지
• /
• 제26권1호
• /
• pp.79-87
• /
• 2021

본 논문에서는 소량의 레이블 데이터로 딥러닝 기반 분류 모델을 훈련할 때 적용되는 준 지도 학습 기법 (semi-supervised learning: SSL)에 대해서 분석한다. 기존의 준 지도 학습 기법은 크게 일관성 정규화 (consistency regularization), 엔트로피 기반 (entropybased), 의사 레이블링 (pseudo labeling)으로 구분할 수 있다. 우선, 각 준 지도 학습 기법의 알고리즘에 대해서 서술한다. 실험에서는 준 지도학습 기법을 레이블 데이터의 수를 변화시키면서 훈련 후 분류 정확도를 평가한다. 최종적으로 실험 결과를 바탕으로 기존 준 지도 학습 기법의 한계에 대해서 서술하고, 분류 성능을 향상하기 위한 연구 방향을 제시한다.

• 센서학회지
• /
• 제24권6호
• /
• pp.368-372
• /
• 2015

In this paper, we proposed a post-processing technique for improving classification performance of electronic nose (E-Nose) system which may be occurred drift signals from sensor array. An adaptive radial basis function network using stochastic gradient (SG) and singular value decomposition (SVD) is applied to process signals from sensor array. Due to drift from sensor's aging and poisoning problems, the final classification results may be showed bias and fluctuations. The predicted classification results with drift are quantized to determine which identification level each class is on. To mitigate sharp fluctuations moving-averaging (MA) technique is applied to quantized identification results. Finally, quantization and some edge correction process are used to decide levels of the fluctuation-smoothed identification results. The proposed technique has been indicated that E-Nose system was shown correct odor identification results even if drift occurred in sensor array. It has been confirmed throughout the experimental works. The enhancements have produced a very robust odor identification capability which can compensate for decision errors induced from drift effects with sensor array in electronic nose system.

• 대한공간정보학회지
• /
• 제4권2호
• /
• pp.181-193
• /
• 1996

본 연구는 원격탐사의 수치화상분류에 적용된 바 없는 정준상관분석(Canonical Correlation Analysis)기법을 무감독분류한 위성화상데이터에 적용하여 토지피복분류하는 새로운 방법을 개발하는 것을 목적으로 한다. 개발된 분류기법은 기존의 분류기법인 최대우도분류기법에 비해 분류기준용 표본데이터 선정이 용이함을 알 수 있었다. 즉, 정준상관분석에 의한 분류결과는 분류기준용 표본데이터의 선정위치에 거의 영향을 받지 않는다. 또한 무감독분류 후 정준상관분석에 의해 결정된 각 군집의 토지피복은 최대우도분류를 위한 사전정보로 활용정보로 활용가능하다. 동일한 분류기준용 표본데이터 사용시, 무감독분류 후 정준상관분석에 의한 분류가 최대우도분류보다 분류정확도가 우수하였다. 이상과 같은 결과로 판단해 볼 때 연구에서는 시도된 분류기법은 원격탐사의 분류기법 분야에서 실용화 될 수 있으며, 나아가서는 GIS 데이터베이스 구축에 중요한 역학을 할 수 있을 것이다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2008년도 International Symposium on Remote Sensing
• /
• pp.29-32
• /
• 2008

In this paper, we propose one unsupervised classification technique using the degree of polarization (DoP) and the co-polarized phase-difference (CPD) statistics, instead of the entropy and alpha. It is shown that the DoP is closely related to the entropy, and the CPD to the alpha. The DoP explains the feature how much the effect of multiple reflections is contained. Hence, the DoP could be used as an important factor for classifying classes. The CPD can also be computed from the measured Mueller matrix elements. For the smooth surface scattering, the CPD is about $0^{\circ}$, and for dihedral-type scattering, the CPD is about $180^{\circ}$. A DoP-CPD diagram with appropriate boundaries between six different classes is developed based on the SAR image. The classification results are compared with the existing Entropy-alpha diagram as well as the IPL-AirSAR polarimetric data. The technique may have capability to classify an SAR image into six major classes; a bare surface, a village, a crown-layer short vegetation canopy, a trunk-layer short vegetation canopy, a crown-layer forest, and a trunk-dominated forest. Based on the DoP and CPD analysis, a simple three-component decomposition technique was also proposed.