• Journal of the Korea Institute of Military Science and Technology
• /
• v.16 no.4
• /
• pp.537-544
• /
• 2013

This paper proposes a recursive localized Normalized Mutual Information(NMI) algorithm to overcome shortcomings of the conventional NMI algorithm and the localized NMI algorithm which proposed before. The localized NMI algorithm divides images into few fixed size segments and applies NMI algorithm to each segments. By extension, the proposed algorithm uses variable size segments using its characteristic. Dividing each segment recursively, the algorithm selects a suitable segment size and improves a performance of the image registration. Experimental result shows the performance of the proposed algorithm.

• International Journal of Control, Automation, and Systems
• /
• v.3 no.2
• /
• pp.143-151
• /
• 2005

This paper considers a robust decentralized $H_\infty$ control problem for uncertain large-scale interconnected systems. The uncertainties are assumed to be time-invariant, norm-bounded, and exist in subsystems. A design method based on the bounded real lemma is developed for a dynamic output feedback controller, which is reduced to a feasibility problem for a nonlinear matrix inequality (NMI). It is proposed to solve the NMI iteratively by the idea of homotopy, where some of the variables are fixed alternately on each iteration to reduce the NMI to a linear matrix inequality (LMI). A decentralized controller for the nominal system is computed first by imposing structural constraints on the coefficient matrices gradually. Then, the decentralized controller is modified again gradually to cope with the uncertainties. A given example shows the efficiency of this method.

• Journal of the Korean Applied Science and Technology
• /
• v.19 no.3
• /
• pp.222-244
• /
• 2002

A mixture of methyl ester derivatives of fatty acids from the oils of pine nuts was well resolved to five fractions differing by degree of unsaturation by silver ion solid-phase extraction column chromatography ($Ag^{+}$-SEC). Polyunsaturated fatty acid with non-methylene interrupted conjugated double bond (NMiDB) radical held more strongly to silver ions in the column than methylene interrupted conjugated double bond (MiDB) one when they had the same number of double bonds. Although both the picolinyl ester and DMOX derivative provided clear mass ion species powerful enough to elucidate the structure of the polyunsaturated fatty acid (PUFA) with NMiDB and/or methylene interrupted conjugated double bond (MiDB) radical in the oils, the picolinyl ester of PUFA with NMiDB radical did not provide a cluster of mass ions neighboring diagnostic mass ions induced by the double bond in the proximal to the carboxyl group. However, the DMOX derivative of PUFA with NMiDB group as well as MiDB showed abundant mass ion species differing by gaps of 12 amu, which made it possible with greater ease to locate the double bonds in the molecule. The oil contained $C_{18:2{\omega}6}$ (46.2 %) and $C_{18:1{\omega}9}$ (25.4 %) as main components, and considerable amounts of PUFAs with NMiDB radical such as ${\Delta}^{5.\;9.\;12}-C_{18:3}$ (16.0 %), ${\Delta}^{5.\;9}-C_{18:2}$ (2.3 %) and ${\Delta}^{5.\;11.\;14}-C_{20:3}$ (0.8 %).

• Fisheries and Aquatic Sciences
• /
• v.27 no.5
• /
• pp.276-282
• /
• 2024

The present study was conducted to investigate the behavior and distribution characteristics of fishes near an artificial reef close to the waters of Uljin marine ranch. A 200-kHz, dual-beam frequency transducer was attached to the side of a ship for acoustic measurements. The fish formed small groups in the bottom layer near the artificial reef around the afternoon-sunset period; at night, the fish did not form groups and remained individually scattered. During dawn-sunrise and morning, the fish formed groups again and were found near the upper and middle layers of the artificial reef. High density of fish occurred near the middle of the nautical zone during morning, afternoon-sunset, and dawn-sunrise, periods; at nighttime, the distribution was uniform across the entire zone. Moreover, the mean Nautical Area Scattering Coefficients (NASC, m²/nmi²) value was highest during dawn-sunrise at 400.2 m²/nmi², similar during night and morning (100.5 m²/nmi²), and lowest during afternoon-sunset (20.1 m²/nmi²). The present study is expected to provide the background for understanding the behavioral characteristics of fish living near artificial reefs and estimating the density and biomass of fish.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.54 no.2
• /
• pp.146-156
• /
• 2018

Acoustic surveys were conducted in the seas surround the South Korea (South Sea A, South Sea B (waters around the Jeju Island), West Sea and East Sea) in spring and autumn in 2016. First, the vertical and horizontal distributions of fisheries resources animals were examined. In most cases vertical acoustic biomass was high in surface water and mid-water layers other than South Sea A in autumn and West Sea. The highest vertical acoustic biomass showed at the depth of 70-80 m in the South Sea A in spring ($274.4m^2/nmi^2$) and the lowest one was 10-20 m in the West Sea in autumn ($0.4m^2/nmi^2$). With regard to the horizontal distributions of fisheries resources animals, in the South Sea A, the acoustic biomass was high in eastern and central part of the South Sea and the northeast of Jeju Island ($505.4-4099.1m^2/nmi^2$) in spring while it was high in eastern South Sea and the coastal water of Yeosu in autumn ($1046.9-2958.3m^2/nmi^2$). In the South Sea B, the acoustic biomass was occurred high in the southern and western seas of Jeju Island in spring ($201.0-1444.9m^2/nmi^2$) and in the southern of Jeju Island in autumn ($203.7-1440.9m^2/nmi^2$). On the other hand, the West Sea showed very low acoustic biomass in spring (average NASC of $1.1m^2/nmi^2$), yet high acoustic biomass in the vicinity of 37 N in autumn ($562.6-3764.2m^2/nmi^2$). The East Sea had high acoustic biomass in the coastal seas of Busan, Ulsan and Pohang in spring ($258.7{\sim}976.4m^2/nmi^2$) and of Goseong, Gangneung, Donghae, Pohang and Busan in autumn ($267.3-1196.3m^2/nmi^2$). During survey periods, fish schools were observed only in the South Sea A and the East Sea in spring and the West Sea in autumn. Fish schools in the South Sea A in spring were small size ($333.2{\pm}763.2m^2$) but had a strong $S_V$ ($-49.5{\pm}5.3dB$). In the East Sea, fish schools in spring had low $S_V$ ($-60.5{\pm}14.5dB$) yet had large sizes ($537.9{\pm}1111.5m^2$) and were distributed in the deep water depth ($83.5{\pm}33.5m$). Fish schools in the West Sea in autumn had strong $S_V$ ($-49.6{\pm}7.4dB$) and large sizes ($507.1{\pm}941.8m^2$). It was the first time for three seas surrounded South Korea to be conducted by acoustic surveys to understand the distribution and aggregation characteristics of fisheries resources animals. The results of this study would be beneficially used for planning a future survey combined acoustic method and mid-water trawling, particularly deciding a survey location, a time period, and a targeting water depth.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.165-167
• /
• 2003

To evaluate the thermal and water stress of vegetation canopy in Southern Qu$\'{e}$bec, leaf water status was evaluated from vegetation indices derived from SPOT VEGETATION images and surface temperature from NOAA AVHRR images. This study was conducted by investigating vegetation conditions for two different periods, from June to August, 1999 and 2000. The vegetation indices were integrated for the evaluating vegetation conditions as a new index, normalized moisture index (NMI). A trapezoid was defined by the NMI and surface temperature, and the thermal and water status of the vegetation canopy was determined according to separate small sections within the trapezoid.

• Journal of the Architectural Institute of Korea Structure & Construction
• /
• v.34 no.6
• /
• pp.63-69
• /
• 2018

The machine learning model can capture the dynamics of building systems with less inputs than the first principle based simulation model. The training data for developing a machine learning model are usually selected in a heuristic manner. In this study, the authors developed a machine learning model which can describe supply air temperature from an AHU in a real office building. For rational reduction of the training data, the progressive sampling method was used. It is found that even though the progressive sampling requires far less training data (n=60) than the offline regular sampling (n=1,799), the MBEs of both models are similar (2.6% vs. 5.4%). In addition, for the update of the machine learning model, the normalized mutual information (NMI) was applied. If the NMI between the simulation output and the measured data is less than 0.2, the model has to be updated. By the use of the NMI, the model can perform better prediction ($5.4%{\rightarrow}1.3%$).

• IEIE Transactions on Smart Processing and Computing
• /
• v.4 no.5
• /
• pp.311-317
• /
• 2015

The aim of this study is to develop a 3D registration algorithm for positron emission tomography/computed tomography (PET/CT) and magnetic resonance (MR) images acquired from independent PET/CT and MR imaging systems. Combined PET/CT images provide anatomic and functional information, and MR images have high resolution for soft tissue. With the registration technique, the strengths of each modality image can be combined to achieve higher performance in diagnosis and radiotherapy planning. The proposed method consists of two stages: normalized mutual information (NMI)-based global matching and independent component analysis (ICA)-based refinement. In global matching, the field of view of the CT and MR images are adjusted to the same size in the preprocessing step. Then, the target image is geometrically transformed, and the similarities between the two images are measured with NMI. The optimization step updates the transformation parameters to efficiently find the best matched parameter set. In the refinement stage, ICA planes from the windowed image slices are extracted and the similarity between the images is measured to determine the transformation parameters of the control points. B-spline. based freeform deformation is performed for the geometric transformation. The results show good agreement between PET/CT and MR images.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.50 no.2
• /
• pp.124-130
• /
• 2014

We aimed to assess the behavior characteristics of fish on an artificial reef using hydroacoustic techniques. The acoustical survey was conducted with a 200 kHz dual beam transducer while fishing on the stone combination reef of Uljin. A school of fish were detected on the artificial reef before sunset and floated on the artificial reef at 30 minutes after sunset. The density (Nautical Area Scattering Coefficient, NASC) of fish that floated on the artificial reef after sunset was about $600m^2/nmi^2$; similar observations (about $50m^2/nmi^2$ or less) were noted after 19:00 hours. Fish caught by fishing on the artificial reef were Sebastes schlegeli, Hexagrammos otakii, Sebastes thompsoni, and Conger myriaster. Resultantly, we demonstrated that hydroacoustic techniques are useful for detecting behavior characteristics of fish in the artificial reef. Such results can be used for basic data to estimate the install effect of the artificial reefs.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.6
• /
• pp.37-48
• /
• 2012

Image registration is a process to establish the spatial correspondence between the images of same scene, which are acquired at different view points, at different times, or by different sensors. In this paper, we propose an effective registration method for images acquired by multi-sensors, such as EO (electro-optic) and IR (infrared) sensors. Image registration is achieved by extracting features and finding the correspondence between features in each input images. In the recent research, the multi-sensor image registration method that finds corresponding features by exploiting NMI (Normalized Mutual Information) was proposed. Conventional NMI-based image registration methods assume that the statistical correlation between two images should be global, however images from EO and IR sensors often cannot satisfy this assumption. Therefore the registration performance of conventional method may not be sufficient for some practical applications because of the low accuracy of corresponding feature points. The proposed method improves the accuracy of corresponding feature points by combining the gradient orientation as spatial information along with NMI attributes and provides more accurate and robust registration performance. Representative experimental results prove the effectiveness of the proposed method.