• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.6-10
• /
• 2007

Classification of hyperspectral images is challenging. A very high dimensional input space requires an exponentially large amount of data to adequately and reliably represent the classes in that space. In other words in order to obtain statistically reliable classification results, the number of necessary training samples increases exponentially as the number of spectral bands increases. However, in many situations, acquisition of the large number of training samples for these high-dimensional datasets may not be so easy. This problem can be overcome by using multiple classifiers. In this paper we compared the effectiveness of two approaches for creating multiple classifiers, feature selection and feature extraction. The methods are based on generating multiple feature subsets by running feature selection or feature extraction algorithm several times, each time for discrimination of one of the classes from the rest. A maximum likelihood classifier is applied on each of the obtained feature subsets and finally a combination scheme was used to combine the outputs of individual classifiers. Experimental results show the effectiveness of feature extraction algorithm for generating multiple classifiers.

• Journal of the Korea Computer Industry Society
• /
• v.7 no.5
• /
• pp.473-480
• /
• 2006

Extraction is tile most important step in global minimization. Its approache is to identify and extract subexpressions, which are multiple-cubes or single-cubes, common to two or more expressions which can be used to reduce the total number of literals in a Boolean network. Extraction is described as either algebraic or Boolean according to the trade-off between run-time and optimization. Boolean extraction is capable of providing better results, but difficulty in finding common Boolean divisors arises. In this paper, we present a new method for Boolean extraction to remove the difficulty. The key idea is to identify and extract two-cube Boolean subexpression pairs from each expression in a Boolean network. Experimental results show the improvements in the literal counts over the extraction in SIS for some benchmark circuits.

• Korean Institute of Interior Design Journal
• /
• v.17 no.5
• /
• pp.99-106
• /
• 2008

Based on extraction items derived from previous related studies on viewing experiences in art museums, this study is conducted to investigate extraction factors that affect viewing satisfactions and to suggest a guideline for an effective viewing environment by clarifying a hierarchy among the extraction factors. For this study, a survey was given to museum visitors, and statistical analyses were conducted on data obtained from the survey. The results of this study are summarized as follows; 1. From an analysis of extraction items that affect overall viewing satisfaction, space and art works were found to be relatively satisfactory. 2. From correlation analyses of extraction items, a degree of concentration on art works was found to most 'affect the viewing satisfactions of art museums. 3. From factor analyses, extraction items were reduced to 11 extraction factors, and a simple extraction structure affecting the viewing satisfactions in art museums. 4. From multiple regression analyses, a extraction factors were derived, and a relative hierarchy among the factors was found.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.8
• /
• pp.1201-1206
• /
• 2013

Active Appearance Model(AAM) is one of the facial feature extraction techniques. In this paper, we propose the Multiple Active Appearance Model(MAAM). Proposed method uses two AAMs. Each AAM trains using different training parameters. It causes that each AAM has different strong points. One AAM complements the weak points in the other AAM. We performed the facial feature extraction on the 100 images to verify the performance of MAAM. Experiment results show that MAAM gives more accurate results than AAM with less fitting iteration.

• Journal of Korea Multimedia Society
• /
• v.7 no.4
• /
• pp.467-473
• /
• 2004

Caption in video plays an important role that delivers video content. Existing caption region extraction methods are difficult to extract caption region from background because they are sensitive to noise. This paper proposes the method to extract caption region in sports video using multiple frame merge and MBR(Minimum Bounding Rectangles). As preprocessing, adaptive threshold can be extracted using contrast stretching and Othu Method. Caption frame interval is extracted by multiple frame merge and caption region is efficiently extracted by median filtering, morphological dilation, region labeling, candidate character region filtering, and MBR extraction.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.1
• /
• pp.123-144
• /
• 2014

Wireless sensor networks are generally deployed for specific applications to accomplish certain objectives over a period of time. To fulfill these objectives, it is crucial that the sensor network continues to function for a long time, even if some of its nodes become faulty. Energy efficiency and fault tolerance are undoubtedly the most crucial requirements for the design of an information extraction protocol for any sensor network application. However, most existing software agent based information extraction protocols are incapable of satisfying these requirements because of static agent itineraries and large agent sizes. This paper proposes an Information Extraction protocol based on Multiple software Agents with Dynamic Itineraries (IEMADI), where multiple software agents are dispatched in parallel to perform tasks based on the query assigned to them. IEMADI decides the itinerary for an agent dynamically at each hop using local information. Through mathematical analysis and simulation, we compare the performance of IEMADI with a well known static itinerary based protocol with respect to energy consumption and response time. The results show that IEMADI provides better performance than the static itinerary based protocols.

• Mass Spectrometry Letters
• /
• v.5 no.3
• /
• pp.79-83
• /
• 2014

Three different dutasteride extraction methods were compared and a method based on liquid-liquid extraction (LLE) using methyl tert-butyl ether and methylene chloride was proved to be more effective than others for the extraction of dutasteride and finasteride, the internal standard (IS), from rat plasma. Additionally, a method composed of the LLE extraction, liquid chromatography, and multiple reaction monitoring (MRM) to target dutasteride and IS was validated by assessing specificity, linearity ($r^2$ = 0.9993, 5 - 400 ng/mL), sensitivity (the limit of detection: 4.03 ng/mL; the limit of quantitation: 12.10 ng/mL), accuracy (intra-day: 89.4 - 105.9%; inter-day: 84.9 - 100.9%), precision (intra-day: 0.8 - 6.9%; inter-day: 2.9 - 15.9%), and recovery (84.7 - 107.8%). Since the validated method was successfully applied to a pharmacokinetic study of dutasteride, it can be useful for the pharmacokinetic evaluation of newly developed dutasteride formulations.

• Ocean Systems Engineering
• /
• v.2 no.2
• /
• pp.137-145
• /
• 2012

Recent developments in underwater image recognition methods have received large attention by the ocean engineering researchers. In this paper, an improved bi-dimensional empirical mode decomposition (BEMD) approach is employed to decompose the given underwater image into intrinsic mode functions (IMFs) and residual. We developed a joint algorithm based on BEMD and Canny operator to extract multi-pixel edge features at multiple scales in IMFs sub-images. So the multiple pixel edge extraction is an advantage of our approach; the other contribution of this method is the realization of the bi-dimensional sifting process, which is realized utilizing regional-based operators to detect local extreme points and constructing radial basis function for curve surface interpolation. The performance of the multi-pixel edge extraction algorithm for processing underwater image is demonstrated in the contrast experiment with both the proposed method and the phase congruency edge detection.

• Journal of the Institute of Convergence Signal Processing
• /
• v.12 no.3
• /
• pp.176-180
• /
• 2011

Even though algorithms for watermark embedding and extraction step are important issue for digital watermarking, watermark selection and post-processing can give us an opportunity to improve our algorithms and achieve higher performance. For this reason, we summarized the possibilities of improvements for digital watermarking by referring to the watermark merging techniques rather than embedding and extraction algorithms in this paper. We chose Cox's function as main embedding and extraction algorithm, and multiple barcode watermarks as a watermark. Each bit of the multiple copies of barcode watermark was embedded into a gray-scale image with Cox's embedding function. After extracting the numbers of watermark, we applied the watermark merging techniques; including the simple merging, N-step iterated merging, recover merging and combination of iterated-recover merging. Main consequence of our paper was the fact of finding out how multiple barcode watermarks and merging techniques can give us opportunities to improve the performance of algorithm.

• Smart Media Journal
• /
• v.7 no.4
• /
• pp.79-89
• /
• 2018

In recent years, falls among elderly people have gained serious attention as a major cause of injuries. Falls often lead to fatal consequences due to lack of prompt response and rescue. Therefore, a more accurate fall detection system and an effective feature extraction technique are required to prevent and reduce the risk of such incidents. In this paper, we proposed an efficient feature extraction technique based on multiple sliding windows and validated it through a series of experiments using supervised learning algorithms. The experiments were conducted using the public datasets obtained from tri-axial accelerometers. The results depicted that extraction of the feature from adjacent sliding windows led to high accuracy in supervised machine learning-based fall detection. Also, the experiments conducted in this study suggested that the best accuracy can be achieved by keeping the window size as small as 2 seconds. With the kNN classifier and dataset from wearable sensors, the experiments achieved accuracy rates of 94%.