• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.979-995
• /
• 2023

As wildfires are difficult to predict, real-time monitoring is crucial for a timely response. Geostationary satellite images are very useful for active fire detection because they can monitor a vast area with high temporal resolution (e.g., 2 min). Existing satellite-based active fire detection algorithms detect thermal outliers using threshold values based on the statistical analysis of brightness temperature. However, the difficulty in establishing suitable thresholds for such threshold-based methods hinders their ability to detect fires with low intensity and achieve generalized performance. In light of these challenges, machine learning has emerged as a potential-solution. Until now, relatively simple techniques such as random forest, Vanilla convolutional neural network (CNN), and U-net have been applied for active fire detection. Therefore, this study proposed an active fire detection algorithm using state-of-the-art (SOTA) deep learning techniques using data from the Advanced Himawari Imager and evaluated it over East Asia and Australia. The SOTA model was developed by applying EfficientNet and lion optimizer, and the results were compared with the model using the Vanilla CNN structure. EfficientNet outperformed CNN with F1-scores of 0.88 and 0.83 in East Asia and Australia, respectively. The performance was better after using weighted loss, equal sampling, and image augmentation techniques to fix data imbalance issues compared to before the techniques were used, resulting in F1-scores of 0.92 in East Asia and 0.84 in Australia. It is anticipated that timely responses facilitated by the SOTA deep learning-based approach for active fire detection will effectively mitigate the damage caused by wildfires.

• Journal of KIISE:Information Networking
• /
• v.35 no.5
• /
• pp.409-414
• /
• 2008

In mobile ad hoc networks, most of on demand routing protocols such as DSR and AODV do not deal with traffic load during the route discovery procedure. To solve the congestion and achieve load balancing, many protocols have been proposed. However, the existing load balancing schemes has only considered avoiding the congested route in the route discovery procedure or finding an alternative route path during a communication session. To mitigate this problem, we have proposed a new scheme which considers the packet bursting mechanism in congested nodes. The proposed packet bursting scheme, which is originally introduced in IEEE 802.11e QoS specification, is to transmit multiple packets right after channel acquisition. Thus, congested nodes can forward buffered packets promptly and minimize bottleneck situation. Each node begins to transmit packets in normal mode whenever its congested status is dissolved. We also propose two threshold values to define exact overloaded status adaptively; one is interface queue length and the other is buffer occupancy time. Through an experimental simulation study, we have compared and contrasted our protocol with normal on demand routing protocols and showed that the proposed scheme is more efficient and effective especially when network traffic is heavily loaded.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.11 no.2
• /
• pp.145-152
• /
• 2001

This study was performed to evaluate the coke oven emissions (COE) and polynuclear aromatic hydrocarbon levels in coke manu-facturing industry, secondary lead smelting industry and glass bottle manufacturing industry. 1. There were no significant difference between the means of personal samples and area samples by the types of industry(p>0.05). The levels of airborne total particulates of the secondary lead smelting industry was the highest($2.30mg/m^3$), and those of the coke manu-facturing industry and glass bottle manu facturing industry were $1.95mg/m^3$ and $1.37mg/m^3$. The concentration of COE was the highest in the glass bottle manufacturing industry($0.79mg/m^3$), and in order of $0.19mg/m^3$ in the coke manufacturing industry and $0.06mg/m^3$ in the secondary lead smelting industry. COE/total particulates(%) was highest in the glass bottle manufacturing industry(58.1%) and in order of 10.3% in the coke manufacturing industry and 3.1% in secondary lead smelting industry. There were significant differences in the total particle concentration and COE by the types of industry(p<0.05). 2. The levels of airborne total particulates was the highest at the smelting process of secondary lead smelting industry($2.30{\pm}0.72mg/m^3$), and the lowest at the smelting process of glass bottle manufacturing industry ($0.99{\pm}1.22mg/m^3$) Concentration of COE was the highest at the casting process of glass bottle manufacturing industry ($1.09{\pm}1.15mg/m^3$), the lowest at the smelting process of secondary lead smelting industry ($0.06{\pm}0.03mg/m^3$). The COE/total particulates(%) was the highest at the casting process of glass bottle manufacturing industry($65.9{\pm}20.5%$), and the lowest at the smelting process of secondary lead smelting indusry($3.1{\pm}2.7%$). 3. There were positive correlations between level of The airborne total particulates and concentration of COE in coke manufacturing industry and glass bottle manufacturing industry (p<0.05), but negative correlation in secondary lead smelting industry. 4. The numbers of case and rates that over the Threshold Limit Values(TLVs) were 24 (77.4%)cases in glass bottle manufacture, 14(23.7%) cases in the coke manufacturing industry and no one case in secondary lead smelting industry. Total numbers of case and rates that over TLVs were 38( 35.5%) cases. 5. The limit of detection(LOD) for PAH was $10{\mu}g/ml$ in standard sample. All PAH levels of the cokes manufacturing industry and the secondary lead smelting industry and the glass bottle manufacturing industry were trace or not to detect.

• Journal of Korean Society for Geospatial Information Science
• /
• v.22 no.4
• /
• pp.135-142
• /
• 2014

As interest in social safety has recently increased at the national level, the various activities which can effectively prevent crimes are being carried out. Because the existing maps related to crimes provide the information about the present condition of crimes by administrative district for users, women and pedestrians who go by night could not actually grasp safe roads in advance. Therefore, this study developed the methodology that can easily extract dangerous areas due to crimes by the digital map 2.0. In the digital map 2.0, location and attribute information of center-lines of roads and building layers were used to find dangerous areas of crimes in these layers. Pavement materials and road width which are already built by the attribute information were used in the center-lines of roads. Crossing angles that roads and roads cross each other were additionally extracted and utilized. The attribute information about building types were input in the building layers of the digital map 2.0. The areas that are more the threshold values set by totaling up all the risk scores when considering pavement materials, road width, crossing angles of road, and building types in the center-lines of roads and road crossings were extracted as the dangerous areas that crimes can occur. Verification of the developed methodology was done by experiment. In the spatial apsect, the dangerous areas of crimes could be found by using the digital 2.0, roads, and building layers only through the experiment. In the administrative aspect to prevent crimes, additional installation of safety facilities such as street lights and security lights in the identified areas which are vulnerable for crimes is thought to be increasing safety of dangerous areas.

• Journal of fish pathology
• /
• v.26 no.2
• /
• pp.65-75
• /
• 2013

Population growth equation of scuticociliate Miamiensis avidus was obtained from the experimental results of in vitro culture condition to estimate the growth rate and carrying capacity from the growth equation. In addition, intraperitoneal infections into olive flounder Paralichthys olivaceus were carried out into 2 different conditions: different concentrations of M. avidus in same water temperature and same concentration of M. avidus in different water temperatures. Olive flounder mortality was threshold dependent with both the temperature and M. avidus density parameters. In this paper, we propose a mathematical model to study M. avidus growth in olive flounder based upon the interactions between parasite and host. The mathematical model was logistic growth differential equation (1.2). The parameters were found with Matlab program through the Levenberge-Marquardt method. In theorem, equilibrium values between the infected fish population and dead population could found. Our equilibrium points were a stable equilibrium and an unstable equilibrium. From the equation (1.6), it was possible to predict the amount of cumulative mortality of olive flounder along with the time after M. avidus infection.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.15 no.4
• /
• pp.199-203
• /
• 2015

With the development of industry these days, the demand for electric power increases and the larger capacity for power transfer is required. The scales of facilities should become larger; and the relative systems are required to operate with a higher degree of reliability. Therefore, stabilization of electric power systems is an important issue. The high degree of reliability required in the process of production and supply of electric power is an essential part of industrial society. Accident such as blackouts causes a hugh amount of economic losses to the high-tech industrial society dependent upon electric power. This paper is about the basic study of the relations between the load current and lifetime of power cables in operation. In order to do the research, we installed a current transformer and an equipment for measuring temperature at the 6.6. kV cables in operation. The two equipments have been installed on the cable systems in operation for the last 20 years. Since the insulation resistance of most of the cables showed the value larger than the threshold, it was not easy to tell the remaining lifetime of cables. The load current of the cables was almost constant. The surrunding temperature was $15{\sim}25^{\circ}C$, little variation of temperature values.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.20 no.1
• /
• pp.135-143
• /
• 2018

The effect of elevated temperature on temperate paddy rice will be significant for dependable food supply in East Asia. Using temperature gradient field chamber (TGFC), which was designed to make the horizontal air temperature gradient by $0^{\circ}C$ to $3^{\circ}C$ higher than outside, we examined the measurement to understand the effects of extremely high temperature on paddy rice. In particular, the data of the year 2016, the worst heat wave in over 22 years, was analyzed in this study. The rice height in the relatively warmed condition was rapidly increased during early growth stage. However, the average grain weight and number of spikelet per panicle in the warmed chamber condition were gradually declined with increasing air temperature averaged for 40 days after first heading in each chamber. In particular, the grain yield was more dramatically decreased by the raising temperature because the percent ripened grain was quickly dropped as getting over the threshold temperature for pollination. Therefore, the surplus photosynthetic product by such lower grain filling rate may disturbed the decreases of the NDVI (Normalized Difference Vegetation Index) and SPAD chlorophyll values after first (normal) heading. In addition, the late-emerging head grain were appeared. However, this yield was too small to recover the normal yields decreased by extremely high temperature condition. Our result represented that the warmed condition in 2016 would be the critical limit for the stable yield of temperate paddy rice.

• Journal of KIISE:Software and Applications
• /
• v.31 no.3
• /
• pp.379-386
• /
• 2004

Video segmentation is a fundamental task in video indexing and it includes two kinds of shot change detections such as the abrupt transition and the gradual transition. The abrupt shot boundaries are detected by computing the image-based distance between adjacent frames and comparing this distance with a pre-determined threshold value. However, the gradual shot boundaries are difficult to detect with this approach. To overcome this difficulty, we propose the method that detects gradual transition in the MPEG compressed video using the HMM (Hidden Markov Model). We take two different HMMs such as a discrete HMM and a continuous HMM with a Gaussian mixture model. As image features for HMM's observations, we use two distinct features such as the difference of histogram of DC images between two adjacent frames and the difference of each individual macroblock's deviations at the corresponding macroblock's between two adjacent frames, where deviation means an arithmetic difference of each macroblock's DC value from the mean of DC values in the given frame. Furthermore, we obtain the DC sequences of P and B frame by the first order approximation for a fast and effective computation. Experiment results show that we obtain the best detection and classification performance of gradual transitions when a continuous HMM with one Gaussian model is taken and two image features are used together.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.37 no.2
• /
• pp.50-59
• /
• 2000

In this paper, a new preprocessing method is proposed to speedup the watershed-based image segmentation In the proposed method, the gradient correction values of ramp edges are calculated from the positions and width of the ramp edges using Laplacian operator, and then, unlike the conventional method in which the monoscale or multi scale gradient image is directly used as a reference iImage, the reference image is obtained by adding the threshold value to each position of the ramp edges in the monoscale gradient image And the marker image is reconstructed on the reference image by erosion By preprocessing the image for the watershed transformation in such a manner, we can reduce the oversegmentations far more than those of applying the conventional morphological filter to the simple monoscale or multiscale gradient-based reference image Thus, we can reduce the total image segmentation time by reducing the time of postprocessing of region merging, which consumes most of the processing time In the watershed-based image segmentation, Experimental results indicate that the proposed method can speedup the total image segmentation about twice than those of the conventional methods, without the loss of ramp edges and principal edges around the dense-edge region.

• Journal of the Korean Vacuum Society
• /
• v.10 no.3
• /
• pp.289-297
• /
• 2001

Highly-porous $SnO_2$thin films were prepared for recognizing and detecting of the inflammable gases, like butane, propane, LPG, carbon monoxide. To obtain sensing films, Sn, Pt/Sn, Au/Sn, and Pt,Au/Sn films were deposited employing a thermal evaporator for Sn film and a sputter for novel metals of Pt or/and Au. These films were annealed for 2 h at $700^{\circ}C$ to form $SnO_2$-based thin films. The films showed the tetragonal structure and also exhibited many defects and porosity, which could give high sensitivity to thin films. The thin films showed high sensitivity and reproductivity to the tested gases(butane, propane, LPG, and carbon monoxide) to even to low gas concentrations in range of workplace environmental standards. Especially, Pt/$SnO_2$film showed the highest sensitivity to butane, LPG, and carbon monoxide. And pure $SnO_2$ film manifested the highest sensitivity to propane. By using the sensing patterns from the films, we could reliably recognize the kinds and the quantities of the tested inflammable gases within the range of the threshold limit values(TLV) and the lower explosion limit(LEL) through the principal component analysis(PCA).