• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.11-20
• /
• 2006

This paper describes the technique which utilizes a fuzzy neural network to sketch feature extraction in digital images. We configure an artificial neural network and make it learn fuzzy membership functions to decide a local threshold applying to sketch feature extraction. To do this. we put the learning data which is membership functions generated based on optimal feature map of a few standard images into the artificial neural network. The proposed technique extracts sketch features in an images very effectively and rapidly because the input fuzzy variable have some desirable characteristics for feature extraction such as dependency of local intensity and excellent performance and the proposed fuzzy neural network is learned from their membership functions, We show that the fuzzy neural network has a good performance in extracting sketch features without human intervention.

• The Korean Journal of Applied Statistics
• /
• v.24 no.1
• /
• pp.185-191
• /
• 2011

In the credit evaluation study with the assumption of mixture distributions, the ROC curve is a useful method to explore the discriminatory power of default and non-default borrowers. The AROC curve is an adjusted ROC curve that can be identified with the corresponding score and is mathematically analyzed in this work. We obtain patterns of this curve by applying normal distributions. Moreover, the relationship between the AROC curve and many classification accuracy statistics are explored to find the optimal threshold. In the case of equivalent variances of two distributions, we obtain that the local minimum of the AROC curve is estimated at the optimal threshold to maximize certain classification accuracies.

• Atmosphere
• /
• v.19 no.4
• /
• pp.309-318
• /
• 2009

Vulnerability to heat was examined for populations of 6 major cities in South Korea (Seoul, Incheon, Daejeon, Gwangju, Daegu, and Busan). Daily excess mortality and maximum temperature from 1991 to 2005 were employed in this study. The results show that the standardized mortality increase associated with a $1^{\circ}C$ increase in daily maximum temperature above the city-specific threshold explains the heat acclimatization effect better than the threshold temperature itself. The estimated increase in mortality (standardized per 10 million population) associated with a $1^{\circ}C$ increase in temperature above the threshold is 4.8 in Incheon, 4.7 in Seoul, 4.3 in Daejeon, 2.8 in Gwangju, 2.4 in Daegu, and 1.5 in Busan, well reflecting the latitudinal locations and local climates of each city. Climate models project more frequent, more intense, and longer lasting heat waves in most land areas in both hemispheres in the 21st century under increasing greenhouse gas concentrations. In order to mitigate the adverse human health impacts due to excess heat, more detailed characteristics of acclimatization to heat need to be understood and quantified.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.2
• /
• pp.187-195
• /
• 1992

This paper propose on the determination of threshold values for extracting shape information of the objects. First, surface curvatures such as mean curvature and gaussian curvature is calculated from given range data. And then local surface regions are classified into the one of 8 primitives by using the sign of mean curvature H and gaussian curvature K. Also from the statistical viewpoint. the range of the zero of H and K in the range is obtained through the analysis of the relation between mean curvature and gaussian curvature. Finally, the effectiveness of the proposed mithod in this paper is demonstrated by comparing with a case, where the zero threshold is arbitrarily obtained.

• Asia-Pacific Journal of Atmospheric Sciences
• /
• v.54 no.4
• /
• pp.527-544
• /
• 2018

This paper presents a nighttime sea fog detection algorithm incorporating unsupervised learning technique. The algorithm is based on data sets that combine brightness temperatures from the $3.7{\mu}m$ and $10.8{\mu}m$ channels of the meteorological imager (MI) onboard the Communication, Ocean and Meteorological Satellite (COMS), with sea surface temperature from the Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA). Previous algorithms generally employed threshold values including the brightness temperature difference between the near infrared and infrared. The threshold values were previously determined from climatological analysis or model simulation. Although this method using predetermined thresholds is very simple and effective in detecting low cloud, it has difficulty in distinguishing fog from stratus because they share similar characteristics of particle size and altitude. In order to improve this, the unsupervised learning approach, which allows a more effective interpretation from the insufficient information, has been utilized. The unsupervised learning method employed in this paper is the expectation-maximization (EM) algorithm that is widely used in incomplete data problems. It identifies distinguishing features of the data by organizing and optimizing the data. This allows for the application of optimal threshold values for fog detection by considering the characteristics of a specific domain. The algorithm has been evaluated using the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) vertical profile products, which showed promising results within a local domain with probability of detection (POD) of 0.753 and critical success index (CSI) of 0.477, respectively.

• Journal of information and communication convergence engineering
• /
• v.9 no.5
• /
• pp.487-490
• /
• 2011

This paper presents splitting and merging algorithm based on adaptive thresholding. The algorithm first divides the image into blocks, and then compares each block using the calculated thresholding value. The blocks which are same are merged using the certain threshold value and different blocks are split unless it satisfies the threshold value. When the block has been merged, maximum and minimum block sizes are determined then the average block size is determined. After the average block size is determined the average intensity and standard deviation of average block is calculated. The process of thresholding is applied to binarize the image. Finally, the experimental results show that the proposed method distinguishes clearly the background with text in the document image.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.3
• /
• pp.645-654
• /
• 2009

In this study, we present an efficient ambulatory speech audiometric system to detect one's hearing problems at an earlier stage as possible without his or her visit to the audiometric testing facility such in a hospital or a clinic. To estimate a person's hearing threshold level in terms of speech sound response in his or her local environment, a digital assistant(PDA) device is used to generate the speech sound with implementing audiometric Graphic User Interface(GUI) system. Furthermore, a supra-aural earphone is used to measure a subject's hearing threshold level in terms of speech sound by the compensating the transducer's gain by adopting speech sound calibration system.

• Proceedings of the KOSOMBE Conference
• /
• v.1992 no.05
• /
• pp.98-101
• /
• 1992

A new algorithm using entropy operater with variable threshold for edge detection from 2D short axis echocardiogram is proposed. This algorithm is based on brightness, mean value of entropy, and variance value of entropy in local window. This algorithm is effective to process complex echocardiographic images due to the speckle noises, echo dropouts and characteristics of 2D echocardiographic constituents. Results of computer simulation of the proposed algorithm show excellent edge detection performance comparing wi th other edge operators which have been applied to 2D echocardiograms.

• Proceedings of the IEEK Conference
• /
• 1998.06a
• /
• pp.835-838
• /
• 1998

본 연구에서는 칼라 CCD 센서를 제어하여, shading과 .gamma. correction 된 데이터를 읽어 들여, 이를 이진레벨 데이터로 바꾼후, 원래의 다치레벨 또는 이진레벨 데이터를 SCSI나 DMA I/F를 통해 전달하는 ASIC을 설계하였다. 본 ASIC에서는 이진화를 위하여 문자 모드에서는 simple threshold와 LAT(local adaptive threshold) 알고리즘을, 그림모드에서는 stucki error diffusion 알고리즘을 적용하였다. 그리고, 구성은 CCD센서 제어블락, 스텝 모타 제어제어블락, 이미지 축소블락, 데이터 이진화 블락, 그리고 DATA I/F 블락 등으로 이루어져 있다. 또한 사용된 technology는 삼성 0.5um CMOS standard cell이며, 크기는 45K gates(내부 메모리 제외)이고, 160QFP package로 구현되었다. ㅎㅁㅅㄷㄴ (soqn apa

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.4
• /
• pp.468-472
• /
• 2004

This paper presents an intelligent iris recognition system which consists of quality check, iris localization, feature extraction, and verification. For the quality check, the local statistics on the pupil boundary is used. Gaussian mixture model is used to segment and localized the iris region. The feature extraction method is based on an optimal waveform simplification. For the verification, we use an intelligent variable threshold.