• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.139-156
• /
• 2021

The biggest reason for using a deep learning model in image classification is that it is possible to consider the relationship between each region by extracting each region's features from the overall information of the image. However, the CNN model may not be suitable for emotional image data without the image's regional features. To solve the difficulty of classifying emotion images, many researchers each year propose a CNN-based architecture suitable for emotion images. Studies on the relationship between color and human emotion were also conducted, and results were derived that different emotions are induced according to color. In studies using deep learning, there have been studies that apply color information to image subtraction classification. The case where the image's color information is additionally used than the case where the classification model is trained with only the image improves the accuracy of classifying image emotions. This study proposes two ways to increase the accuracy by incorporating the result value after the model classifies an image's emotion. Both methods improve accuracy by modifying the result value based on statistics using the color of the picture. When performing the test by finding the two-color combinations most distributed for all training data, the two-color combinations most distributed for each test data image were found. The result values were corrected according to the color combination distribution. This method weights the result value obtained after the model classifies an image's emotion by creating an expression based on the log function and the exponential function. Emotion6, classified into six emotions, and Artphoto classified into eight categories were used for the image data. Densenet169, Mnasnet, Resnet101, Resnet152, and Vgg19 architectures were used for the CNN model, and the performance evaluation was compared before and after applying the two-stage learning to the CNN model. Inspired by color psychology, which deals with the relationship between colors and emotions, when creating a model that classifies an image's sentiment, we studied how to improve accuracy by modifying the result values based on color. Sixteen colors were used: red, orange, yellow, green, blue, indigo, purple, turquoise, pink, magenta, brown, gray, silver, gold, white, and black. It has meaning. Using Scikit-learn's Clustering, the seven colors that are primarily distributed in the image are checked. Then, the RGB coordinate values of the colors from the image are compared with the RGB coordinate values of the 16 colors presented in the above data. That is, it was converted to the closest color. Suppose three or more color combinations are selected. In that case, too many color combinations occur, resulting in a problem in which the distribution is scattered, so a situation fewer influences the result value. Therefore, to solve this problem, two-color combinations were found and weighted to the model. Before training, the most distributed color combinations were found for all training data images. The distribution of color combinations for each class was stored in a Python dictionary format to be used during testing. During the test, the two-color combinations that are most distributed for each test data image are found. After that, we checked how the color combinations were distributed in the training data and corrected the result. We devised several equations to weight the result value from the model based on the extracted color as described above. The data set was randomly divided by 80:20, and the model was verified using 20% of the data as a test set. After splitting the remaining 80% of the data into five divisions to perform 5-fold cross-validation, the model was trained five times using different verification datasets. Finally, the performance was checked using the test dataset that was previously separated. Adam was used as the activation function, and the learning rate was set to 0.01. The training was performed as much as 20 epochs, and if the validation loss value did not decrease during five epochs of learning, the experiment was stopped. Early tapping was set to load the model with the best validation loss value. The classification accuracy was better when the extracted information using color properties was used together than the case using only the CNN architecture.

• Journal of the Korean earth science society
• /
• v.21 no.3
• /
• pp.287-301
• /
• 2000

In order to investigate the factors and processes affecting the vertical distributions of ozone, we analyzed the ozone profile data measured using ozonesonde from 1995 to 1997 at Pohang city, Korea. In the course of our study, we analyzed temporal and spatial distribution characteristics of ozone at four different heights: surface (100m), troposphere (10km), lower stratosphere (20km), and middle stratosphere (30km). Despite its proximity to a local, but major, industrial complex known as Pohang Iron and Steel Co. (POSCO), the concentrations of surface ozone in the study area were comparable to those typically observed from rural and/or unpolluted area. In addition, the findings of relative enhancement of ozone at this height, especially between spring and summer may be accounted for by the prevalence of photochemical reactions during that period of year. The temporal distribution patterns for both 10 and 20km heights were quite compatible despite large differences in their altitudes with such consistency as spring maxima and summer minima. Explanations for these phenomena may be sought by the mixed effects of various processes including: ozone transport across two heights, photochemical reaction, the formation of inversion layer, and so on. However, the temporal distribution pattern for the middle stratosphere (30km) was rather comparable to that of the surface. We also evaluated total ozone concentration of the study area using Brewer spectrophotometer. The total ozone concentration data were compared with those derived by combining the data representing stratospheric layers via Umkehr method. The results of correlation analysis showed that total ozone is negatively correlated with cloud cover but not with such parameter as UV-B. Based on our study, we conclude that areal characteristics of Pohang which represents a typical coastal area may be quite important in explaining the distribution patterns of ozone not only from surface but also from upper atmosphere.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.34 no.1
• /
• pp.118-128
• /
• 2016

The purpose of this study is to present efficient methods of preserving and managing the fish shelter forest in Mulgun-ri on the southern coast of Korea on the basis of its humanistic, sociological and ecological characteristics. The study object is Korean natural monument No. 150, which is presumed to have been forested by descendants of Jeonju Lee Family who settled there, and village rituals are held every October to pray for the peace of the village. The forest is managed by Namhae-gun as a historical and cultural resource as well as its disaster-preventing, economic, and environmental and ecological functions. The linear form of the area is $23,962.6m^2$ and farmland(48.5%) and urbanization area(38.2%) are extensively located in its periphery area. Actual vegetation was sub-classified into three types of land according to use pressure and whether or not damage was done: land where its stratification was formed; land where it was restored, and the land where it was damaged. Plant communities were sub-classified into Aphananthe aspera community(I) and Zelkova serrata community(II) which had a low use pressure; Z. serrata-Chionanthus retusa-A. aspera community(III) and A. aspera-Z. serrata community(IV) which had a high use pressure; and Celtis sinensis-A. aspera community(V) whose underlayer was damaged by use. Fragmentation of the forest is under way and its inside vegetation growth is hampered due to the installation of traffic and resting facilities such as the through roads costal roads, wooden-deck walkways, parking lots, washstands, etc. As a restoration management plan for this, the following were required: an establishment of preferred restoration area; a selection of restoration vegetation species; and an appropriate restoration method. The damaged area($7,868.2m^2$) will have to be set up as the preferred restoration area; seedlings of restored vegetation species should be raised with dominant species within the forest(i.e., Z. serrata, A. aspera, C. sinensis, and C. retusa) as their 'mother trees' for the benefit of for the next-generation forest; and sub-tree and shrub layer should be complementarily planted with 5 and 115 trees(unit $100m^2$) respectively to facilitate the formation of a multi-layered vegetation structure. In addition, resting facilities scattered inside the forest should be demolished; and indiscriminate use of them should be controlled; management and monitoring should be carried out so that the area can be preserved and restored as a deciduous broad-leaved forest.

• Journal of Cadastre & Land InformatiX
• /
• v.48 no.1
• /
• pp.139-151
• /
• 2018

The National Base Map construction consists of the regular correction work of dividing the whole country into two regions and carrying out the modification Plotting by aerial photographs every two years as well as the real time updating work of correcting the major change feature within two weeks by the field survey and the As-Built Drawing. In the case of the Building Layout Drawing of Korea Real estate Administration intelligence System(KRAS) used for real time updating work of the National base map, the coordinate transformation error is included in the positional error when applied to the National Base Map based on the World Geodetic Reference System as the coordinate system based on the Regional Geodetic Reference System. In addition, National Base Map is registered based on the outline(eaves line) of the building in the Digital Topographic Map, and the Cadastral and Architecture are registered based on the building center line. Therefore, the Building Object management standard is inconsistent. In order to investigate the improvement method, the network RTK survey was conducted directly on a location of the Building Layout Drawing of Korea Real estate Administration intelligence System(KRAS) and the problems were analyzed by comparing with the plane plotting position reference in National Base Map. In the case of the general structure with the difference on the Building center line and the eaves line, beside the location information was different also the difference in the ratio of the building object was different between Building center line and the eave. In conclusion, it is necessary to provide the Base data of the double layer of the Building center line and the outline of the building(eaves line) in order to utilize the Building Layout Drawing of Korea Real estate Administration intelligence System(KRAS). In addition, it is necessary to study an organic map update process that can acquire the up-to-dateness and the accuracy at the same time.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.4
• /
• pp.24-38
• /
• 2019

Due to the dramatic spatial changes caused by industrialization, environmental problems such as air pollution and urban heat island phenomenon, etc. are occurring in cities. In this case, the wind corridor, which is a passage through which fresh and cool air generated in forests outside cities move to the downtown, can be used as a spatial planning method for improving urban environmental problems. Cold air is determined by the characteristics of the flow depending on the topography and land use of cities, and based on this, the medium- and long-term plan should be established. Therefore, this study analyzed the flow of cold air at night through the KLAM_21 model in Suwon-si, Gyeonggi-do, to prepare the basic data required to apply the wind corridors. As a result, it turned out that cold air of Suwon-si was mainly generated from Gwanggyo Mountain that is a large mountain area in the north, and flowed into the urbanization promotion area, and about three hours after sunset, cold air flowed into the downtown. By district, the depth, wind speed, and direction of the cold air layer were formed differently according to the characteristics of the topography and land use. In the areas where large forests were adjacent, the flow of cold air was active. There are three main wind corridors where cold air flows to the downtown of Suwon-si, all of which are formed around rivers. Especially, if the connection between rivers and the surrounding green areas is high, the effect of wind corridors is found to be significant. In order to utilize the wind corridors of Suwon-si, based on the results of this study, it is necessary to make climate maps through actual survey and complex analysis of cold air flow and establish mid-to-long-term plans for the conservation and expansion of major wind corridors.

• Korean Journal of Poultry Science
• /
• v.20 no.1
• /
• pp.1-9
• /
• 1993

This study carried out to build up new synthetic egg production lines which had sex-linked gene for feather color sexing and also superior combining ability for producing the best commercial chicks. The closed flock breeding method was utilized to improve the general performances in the first experiment and combining ability for heterosis was tested for new synthetic line in the second experimental year. In order to test for the egg production ability in cross breeds synthetic lines, the crossing of B$\times$4 B$\times$C, two imported strains and two domestic strains as controls were compared for the general performances. There was on difference in mortality, body weight to 56 weeks of age. Sexual maturity was delayed about 10 days by comparing with other reports, except 153 days of the Manina White, but no difference among mating systems in this experiment. The hen housed egg production in B$\times$A, B$\times$C was 186.3, 191.3 respectively and it was better than the other controls, except ISA imported lines. The hen-day egg production of B$\times$A, B$\times$C was better than other controls, with 75.7%, 76.8% respectively. In the average egg weight, the B$\times$C cross breed was highest with 64.5g. As the sex of hatching chicks was identified by difference of feather color, the genetic composition of synthetic lines must be homogenized. The feather color of female chicks was brown and that of male was silver (99%), In conclusion, the egg production ability of B$\times$A, B$\times$C cross breeds was superior to the imported and domestic lines. Therefore, it suggest that the synthetic lines with sex-linked gene might be utilized for improving egg production performances.

• Journal of Animal Environmental Science
• /
• v.8 no.1
• /
• pp.9-16
• /
• 2002

This study was conducted to develope the new solid separating system which can be efficiently and economically removed the solid parts in high pollutants concentration of pig slurry. The pollutants concentration, BOD$_{5}$ , COD and SS of the slurry used in this study was 15,990($\pm$2,389)mg/l, 20,004($\pm$5,512)mg/l and 26,486($\pm$5,935)mg/l, respectively. After removal of solid part in slurry, the pollutants concentration, BOD$_{5}$, COD and SS was change into 5,617($\pm$690)mg/l, 5,553($\pm$633)mg/land 1,456($\pm$341)mg/l, respectively in the Fixed biological membrane tank. The reduction of the pollutants concentration of suspend liquid through membrane will be allowed to greatly improve the water purification by an Activated sludge method. This separating system consisted of a temporary storage, a circulating tank and a Fixed Biological membrane tank. A temporary storage which has a draining system of screw type and an aeration device played a tremendous role in draining the solid by filled an aeration of 0.3 l/min. A Fixed Biological membrane tank of which a styrofoam filled in a 2/3 volume as a Biological media was fixed by a stainless steel net (pore size : 0.5mm) to separate the liquid layer of influx in them. The separating system efficiency factors were the speed of screw motor, cycle number of slurries in a circulating tank and moisture contents of solid effluent through the screw path. Although the pollutants concentration was very variable in temporary storage, the final concentration of $BOD_5$ and SS, except COD of the suspended liquid in a Fixed biological membrane were not different regardless of cycle number of a circulating tank. Moisture contents of effluent from temporary storage was 73% under the speed 1 ppm of screw motor and 62% under the 1/4rpm of it.

• Journal of Animal Science and Technology
• /
• v.49 no.6
• /
• pp.819-828
• /
• 2007

Two experiments were conducted to observe the effects of direct fed microbials on metabolic characteristics in sheep and milking performance in dairy cows. A metabolic trial with four ruminally cannulated sheep(60±6kg) was conducted in a 4×4 Latin square design to investigate the supplementation effects of Saccharomyces cerevisiae, Clostridium butyricum or mixed microbes of S. cerevisiae and C. butyricum on ruminal fermentation characteristics and whole tract digestibility. Sheep were fed 1.25 kg of total mixed ration(TMR, DM basis) supplemented with S. cerevisiae (2.5g/day), C. butyricum (1.0g/day) or its mixture(S. cerevisiae 1.25g/day+C. butyricum 1g/day), twice daily in an equal volume. But control sheep were fed only TMR. A feeding trial with 28 lactating Holstein cattle was also conducted for 12 weeks to investigate the effects of the same microbial supplements as for the metabolic trial on milking performance. The cows were fed the TMR(control), and fed S. cerevisiae(50g/day), C. butyricum(15g/day) or its mixture (S. cerevisiae 25g/day + C. butyricum 7.5g/day) with upper layer dressing method. Total VFA concentration and the digestibility of whole digestive tract in the sheep increased by supplementation of S. cerevisiae, C. butyricum or their combined microbials compare to control group. The proportion of propionic acid at 1h(P<0.039) and 3h(P<0.022) decreased by supplementation of S. cerevisiae while tended to increase acetic acid proportion at the same times. Daily dry matter intake(DMI) was not influenced by the microbial treatments, but milk yield(P<0.031) and feed efficiency(milk yield/DMI, P<0.043) were higher for the cow received C. butyricum than those for other treatments. The milk fat content was higher (P<0.085) when cows fed S. cerevisiae(4.11%) than that fed the control (4.08%), the diets with C. butyricum (3.85%) and the microbial mixture. Based on the results obtained from the current experiments, supplementation of C. butyricum or mixture with S. cerevisiae might be increased milk fat content and milk productivity of lactating daily cows. (Key words:Saccharomyces cerevisiae, Clostridium butyricum, Fermentation characteristics,

• The Journal of Engineering Geology
• /
• v.23 no.3
• /
• pp.201-216
• /
• 2013

Ground penetrating radar (GPR) surveys were conducted in a sand tank model in a laboratory and at an alluvial field site to detect the groundwater table and to investigate the influence of saturation on GPR response in the unsaturated zone. In the sand tank model, the groundwater table and saturation in the sand layer were altered by injecting water, which was then drained by a valve inserted into the bottom of the tank. GPR vertical reflection profile (VRP) data were obtained in the sand tank model for rising and lowering of the groundwater table to estimate the groundwater table and saturation. Results of the lab-scale model provide information on the sensitivity of GPR signals to changes in the water content and in the groundwater table. GPR wave velocities in the vadose zone are controlled mainly by variations in water content (increased travel time is interpreted as an increase in saturation). At the field site, VRP data were collected to a depth of 220 m to estimate the groundwater table at an alluvial site near the Nakdong river at Iryong-ri, Haman-gun, South Korea. Results of the field survey indicate that under saturated conditions, the first reflector of the GPR is indicative of the capillary fringe and not the actual groundwater table. To measure the groundwater table more accurately, we performed a GPR survey using the common mid-point (CMP) method in the vicinity of well-3, and sunk a well to check the groundwater table. The resultant CMP data revealed reflective events from the capillary fringe and groundwater table showing hyperbolic patterns. The normal moveout correction was applied to evaluate the velocity of the GPR, which improved the accuracy of saturation and groundwater table information at depth. The GPR results show that the saturation information, including the groundwater table, is useful in assessing the hydrogeologic properties of the vadose zone in the field.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.44 no.6
• /
• pp.177-187
• /
• 2016

This study started with the question, "Is the cost of landscape construction work in residential areas measured by public enterprises, 'in response to the needs of consumers?" The study analyzed whether the landscape construction expenditure is being introduced at an appropriate ratio according to the value the residents have regarding landscape elements. Following this, research was conducted for the purpose of providing basic data for improving the efficiency of formulating apartment landscape construction costs in the future. This research proceeded based on a questionnaire survey of residents of apartments, and the content of the questionnaire used frequency analysis and descriptive statistics research methods. To take a look at a comparative analysis of value recognition, in particular, a comparative analysis was performed based on the actual input cost based on the ratio of landscape elements by layer. Conclusions were found as follows: First, the degree of interest in the apartment landscape of the tenants was high, and the value of the landscape was high but realistic satisfaction appeared comparatively low. Second, the awareness of residents' values regarding landscape elements appeared to give "plantings" more value than "facilities". Thirdly, as a result of a mutual comparison between the values recognized by the resident regarding landscape elements and the construction input fee, depending on the landscape elements, it appeared that there is a difference in the ratio of up to 52 times from 1.25. Fourth, the fact that there is a difference in the relative proportion of value recognition and inputting construction cost indicates that it is not possible to respond to the needs of tenants during the construction cost development process. It also shows that the utility of inputting construction costs is low. Therefore, a macro-level examination such as reflecting the existing inflation rate is necessary to develop the efficient landscape construction cost of apartment such as the awareness of the value of the residents regarding landscape elements, out of the customary construction cost formulation method based on the microscopic dimensions of the consumer side.