• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.422-424
• /
• 2021

In this paper, it leads the approach of fusing multiple RGB cameras for visual objects recognition based on deep learning with convolution neural network and 3D Light Detection and Ranging (LiDAR) to observe the environment and match into a 3D world in estimating the distance and position in a form of point cloud map. The goal of perception in multiple cameras are to extract the crucial static and dynamic objects around the autonomous vehicle, especially the blind spot which assists the AV to navigate according to the goal. Numerous cameras with object detection might tend slow-going the computer process in real-time. The computer vision convolution neural network algorithm to use for eradicating this problem use must suitable also to the capacity of the hardware. The localization of classified detected objects comes from the bases of a 3D point cloud environment. But first, the LiDAR point cloud data undergo parsing, and the used algorithm is based on the 3D Euclidean clustering method which gives an accurate on localizing the objects. We evaluated the method using our dataset that comes from VLP-16 and multiple cameras and the results show the completion of the method and multi-sensor fusion strategy.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.34-37
• /
• 2005

An efficient and robust neural network-based scheme is introduced in this paper to perform automatic cloud detection. Unlike many existing cloud detection schemes which use thresholding and statistical methods, we used the artificial neural network methods, the multi-layer perceptrons (MLP) with back-propagation algorithm and radial basis function (RBF) networks for cloud detection from Geostationary satellite images. We have used a simple scene (a mixed scene containing only cloud and clear sky). The main results show that the neural networks are able to handle complex atmospheric and meteorological phenomena. The experimental results show that two methods performed well, obtaining a classification accuracy reaching over 90 percent. Moreover, the RBF model is the most effective method for the cloud classification.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.6
• /
• pp.299-305
• /
• 2020

We have developed a scheduler that additionally consider network performance by extending the Kubernetes developed to manage lots of containers in cloud computing nodes. The network delay adapt characteristics of the compute nodes were learned during server operation and the learned results were utilized to develop placement algorithm by considering the existing measurement units, CPU, memory, and volume together, and it was confirmed that the low delay network service was provided through placement algorithm.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.12
• /
• pp.505-518
• /
• 2023

3D point cloud semantic segmentation is a computer vision task that involves dividing the point cloud into different objects and regions by predicting the class label of each point. Existing 3D semantic segmentation models have some limitations in performing sufficient fusion of multi-modal features while ensuring both characteristics of 2D visual features extracted from RGB images and 3D geometric features extracted from point cloud. Therefore, in this paper, we propose MMCA-Net, a novel 3D semantic segmentation model using 2D-3D multi-modal features. The proposed model effectively fuses two heterogeneous 2D visual features and 3D geometric features by using an intermediate fusion strategy and a multi-modal cross attention-based fusion operation. Also, the proposed model extracts context-rich 3D geometric features from input point cloud consisting of irregularly distributed points by adopting PTv2 as 3D geometric encoder. In this paper, we conducted both quantitative and qualitative experiments with the benchmark dataset, ScanNetv2 in order to analyze the performance of the proposed model. In terms of the metric mIoU, the proposed model showed a 9.2% performance improvement over the PTv2 model using only 3D geometric features, and a 12.12% performance improvement over the MVPNet model using 2D-3D multi-modal features. As a result, we proved the effectiveness and usefulness of the proposed model.

• Korean Journal of Remote Sensing
• /
• v.23 no.2
• /
• pp.71-87
• /
• 2007

In this study, 14 heavy snow events in Yeongdong area which are local phenomena are analyzed using MODIS cloud products provided from NASA/GSFC. The clouds of Yeongdong area at observed at specific time by MODIS are classified into A, B, C Types, based on the characteristic of cloud properties: cloud top temperature, cloud optical thickness, Effective Particle Radius, and Cloud Particle Phase. The analysis of relations between cloud properties and precipitation amount for each cloud type show that there are statistically significant correlations between Cloud Optical Thickness and precipitation amount for both A and B type and also significant correlation is found between Cloud Top Temperature and precipitation amount for A type. However, for C type there is not any significant correlations between cloud properties and precipitation amount. A-type clouds are mainly lower stratus clouds with small-size droplet, which may be formed under the low level cold advection derived synoptically in the East sea. B-type clouds are developed cumuliform clouds, which are closely related to the low pressure center developing over the East sea. On the other hand, C-type clouds are likely multi-layer clouds, which make satellite observation difficult due to covering of high clouds over low level clouds directly related with Yeongdong heavy snow. It is, therefore, concluded that MODIS cloud products may be useful except the multi-layer clouds for understanding the mechanism of heavy snow and estimating the precipitation amount from satellite data in the case of Yeongdong heavy snow.

• Current Optics and Photonics
• /
• v.4 no.3
• /
• pp.174-185
• /
• 2020

Multiscattering occurs when a laser transmits into dense atmosphere targets (e.g. fogs, smoke or clouds), which can cause depolarization effects even though the scattering particles are spherical. In addition, multiscattering effects have additional information about microphysical properties of scatterers. Thus, multiscattering can be utilized to study the microphysical properties of the liquid water cloud. In this paper, a Monte Carlo method was used to simulate multi-scattering transmission properties of Lidar signals in the cloud. The results showed the slope of the degree of linear polarization (SLDLP) can be used to invert the extinction coefficient, and then the cloud effective size (CES) and the liquid water content (LWC) may be easily obtained by using the extinction coefficient and saturation of the degree of linear polarization (SADLP). Based on calculation results, a microphysical properties inversion method for a liquid cloud was presented. An innovative multiscattering polarization Lidar (MSPL) system was constructed to measure the LWC and CES of the liquid cloud, and a new method based on the polarization splitting ratio of the Polarization Beam Splitter (PBS) was developed to calibrate the polarization channels of MSPL. By analyzing the typical observation data of MSPL observation in the northern suburbs of Nanjing, China, the LWC and CES of the liquid water cloud were obtained. Comparisons between the results from the MSPL, MODIS and the Microwave radar data showed that, the microphysical properties of liquid cloud could be retrieved by combining our MSPL and the inversion method.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2017.01a
• /
• pp.79-82
• /
• 2017

IoT devices are interconnected in global network with different functionalities and manage the data transfer in cloud computing. IoT devices can be used anytime, anywhere with any device with different applications and protocols. Same devices but different applications according to end user requirements such as sensors and Wi-Fi devices, reusability of these applications can enhance the development process. However, large number of variations in cloud computing make it difficult the features selection in application because of compatibility issues of devices. In this paper we have proposed multi-Software Product Lines (multi-SPLs) approach to manage the variabilities and commonalities of IoT applications and protocols. Feature modeling is used to manage the commonalities and variabilities of SPL. We proposed that multi-SPLs feature model is more appropriate for modeling of IoT applications and protocols.

• Korean Journal of Computational Design and Engineering
• /
• v.8 no.3
• /
• pp.133-140
• /
• 2003

This paper describes a Multi-step Digitizing Method and Reverse Model generation algorithm for improvement of reverse engineering accuracy. Reverse engineering is the process of reproducing computational model by directly extracting geometric information on the physical objects. For the improvement of measuring data accuracy, we propose a multi-step digitizing method. First, measuring cloud-of-point by use of a laser scanning system. Second, gathering digitizing data by a scanning touch probe. Fine digitizing plan generated from coarse surface model directly from the cloud-of-point and it allows CMM more accurate scanning data. Finally in this paper we propose the algorithm of generating NURB surface from more accurate measuring points.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.34 no.6
• /
• pp.355-362
• /
• 2021

Various blast pressure calculation methods have been developed for predicting the explosion pressure of vapor cloud explosions. Empirical methods include the TNT equivalent method, and multi-energy method. The multi-energy method uses a charge strength that considers environmental factors. Although the Kinsella guideline was provided to determine the charge strength, there are limitations such as guidelines related to ignition sources. In this study, we proposed an improved charge strength guideline, by subdividing the ignition source intensity and expanding the type classification through literature analysis. To verify the improved charge strength guideline, and to compare it with the result obtained using the Kinsella guideline, four vapor cloud explosion cases which could be used to estimate the actual blast pressure were investigated. As a result, it was confirmed that the Kinsella guidelines showed an inaccurate, that is, wider pressure than the actual estimated blast pressure. However, the improved charge strength guideline enabled the selection of the intensity of the ignition source, and more subdivided types through the expansion of classification, hence it was possible to calculate the blast pressure relatively close to that of the actual case.

• Journal of The Korean Astronomical Society
• /
• v.38 no.3
• /
• pp.365-370
• /
• 2005

We have made a multi-wavelength study of the X-ray bright giant shell complex 30 Doradus in the Large Magellanic Cloud (LMC). This is the one of the largest H II complexes in the Local Group. The Australia Telescope Compact Array (ATCA) and the Parkes 64-m single dish observations reveal that the distribution and internal motions of H I gas show the effects of fast stellar winds and supernova blasts. The hot emitting gas within the 30 Doradus complex and the entire giant H II complex are encompassed by an expanding H I shell. We investigate the dynamical age of this H I shell and compare to the age of starbursts occurred in the 30 Doradus nebula using the radiative transfer model and the infrared properties.