• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.6
• /
• pp.834-841
• /
• 2022

Recently, applying artificial intelligence technologies in various fields of production has drawn an upsurge of research interest due to the increase for smart factory and artificial intelligence technologies. A great deal of effort is being made to introduce artificial intelligence algorithms into the defect detection task. Particularly, detection of defects on the surface of metal has a higher level of research interest compared to other materials (wood, plastics, fibers, etc.). In this paper, we compare and analyze the speed and performance of defect classification by combining machine learning techniques (Support Vector Machine, Softmax Regression, Decision Tree) with dimensionality reduction algorithms (Principal Component Analysis, AutoEncoders) and two convolutional neural networks (proposed method, ResNet). To validate and compare the performance and speed of the algorithms, we have adopted two datasets ((i) public dataset, (ii) actual dataset), and on the basis of the results, the most efficient algorithm is determined.

• Journal of Forest and Environmental Science
• /
• v.39 no.2
• /
• pp.73-80
• /
• 2023

Little is known of the life history of Garcinia kola; the objective of this study, therefore, was to assess the fruiting age and tree size of the species in a coastal humid tropical climate plantation condition. A total 103 trees were used in the study viz; 80 ten-year-old trees at reproductive maturity onset and 13 thirty-year-old trees with several cycles of reproduction that constitute two independent variables. Data collected were age of onset of flowering and size at reproductive maturity onset. Relative size at reproductive maturity onset (RSOM) was estimated as size at reproductive maturity onset (SOM) divided by asymptotic maximal size (AMS). Data analysis was conducted using pairwise t-test and principal component analysis (PCA). Reproductive maturity onset (flowering) was recorded in the ten-year-old stand eight (8) years after planting. Mean size at reproductive maturity onset (SOM) was height 5.32±1.7 m, dbh 0.11±0.03 m, total number of branches was 29.6±7.3, crown depth 5.24±1.05 m, crown diameter was 4.78±0.7 m, branch diameter 0.098±0.01 m, leaf length 0.13±0.02 m, leaf breadth 0.37±0.01 m, twig length 0.35±0.11 m and leaf per twig 6±0.84 and asymptotic maximal size (AMS) was height 19.85±0.76 m, dbh 0.95±0.09 m, total number of branches 62±5, crown depth 18.83±0.7 m, crown diameter 12.5±1.64 m, branch diameter 0.5±1.6 m, leaf length 0.16±0.023 m, leaf breadth 0.45±0.12 m, twig length 0.37±0.11 m and leaf per twig 19±7.5. Pairwise t-test analysis showed there was significant differences between SOM and AMS in all growth factors except leaf length, leaf breadth, and twig length. Highest relative size at reproductive maturity onset (RSOM) was recorded in leaf length 0.82, twig length 0.82, and leaf breadth 0.80, while, the lowest was branch diameter 0.11. Four components out of the total of eleven were extracted to explain the relationship in RSOM: Principal component one (PC1) explained 37.23%; PC2 26.4%, PC3 22.73%, and PC4 13.64%.

• Journal of Korean Society on Water Environment
• /
• v.28 no.1
• /
• pp.84-93
• /
• 2012

The study was conducted to analyze the spatio-temporal changes in water quality of the major 36 sampling stations of Nakdong River, depending on each station, season using the 17 water quality variables from 2000 to 2010. The result was verified to interpret the characteristics of water quality variables in a more accurate manners. According to the Principal component analysis (PCA) and Exploratory factor analysis (EFA) results; the results of these analyses were identified 4 factors, Factor 1 (nutrients) included the concentrations of T-N, T-P, $NO_{3}-N$, $PO_{4}-P$, DTN, DTP for sampling station and season, Factor 2 (organic pollutants) included the concentrations of BOD, COD, Chl-a, Factor 3 (microbes) included the concentrations of F.Coli, T.Coli, and Factor 4 (others) included the concentrations of pH, DO. The results of a Cluster analysis indicated that Geumhogang 6 was the most contaminated site, while tributaries and most of the down stream sites of Nakdong River were mainly affected by each nutrients (Factor 1) and organic pollutants (Factor 2). The verification consequence of Confirmatory factor analysis (CFA) from Exploratory factor analysis (EFA) result can be summarized as follows: we could find additional relations between variables besides the structure from EFA, which we obtained through the second-order final modeling adopted in CFA. Nutrients had the biggest impact on water pollution for each sampling station and season. In particular, It was analyzed that P-series pollutant should be controlled during spring and winter and N-series pollutant should be controlled during summer and fall.

• Korean Journal of Food Science and Technology
• /
• v.34 no.1
• /
• pp.13-17
• /
• 2002

The study was to analyze aroma patterns of Gastrodiae Rhizoma by the electronic nose with conducing polymer 32 sensors. Response by the electronic nose was analyzed by the principal component analysis (PCA). Sensory evaluation for the organoleptic characteristics of Gastrodiae Rhizoma was also performed. Raw Gastrodiae Rhizoma was very intensive in overall odor and taste. Hot air-dried $60^{\circ}C$ Gastrodiae Rhizoma was relatively weak in overall odor and taste. Intensity of aroma in the electronic nose was the highest in $60^{\circ}C$ hot air-dried Gastrodiae Rhizoma. As quality factor (QF) calcuated from PCA map of normalized pattern data by thirty two sensors showed less than 2, and so aroma pattern among raw, freezed-dried, and hot-air dried Gastrodiae Rhizoma had no difference. When PCA was performed for normalized pattern data by the selected sensitive ten sensors, QF value between raw and $60^{\circ}C$ hot air-dried Gastrodiae Rhizoma was 2.366. Thus aroma pattern of raw and $60^{\circ}C$ hot air-dried Gastrodiae Rhizoma was discriminated in electronic nose.

• Food Science and Preservation
• /
• v.20 no.3
• /
• pp.348-355
• /
• 2013

This study was conducted to analyze the sensory profiles of commercial rice cookies and snacks in the market and to provide fundamental data on the development of rice cookies and snacks. The sensory characteristics of 10 kinds of Korean commercial rice cookies and snacks were evaluated using a sensory test and were analyzed via quantitative description analysis (QDA) and principal component analysis (PCA), depending on their rice contents and processing methods. The rice cookies and snacks with less than 30% rice contents and that were fried were preferred, and their grainy attribute was related to the preference for them. Then 27 kinds of commercial rice cookies and snacks from three countries, Japan, the U.S.A. and China, were also evaluated and compared with those from Korea. While the Korean commercial rice cookies and snacks were sweet, grainy and crispy, those of Japan were not sweet and instead, were salty and crispy.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.19 no.5
• /
• pp.439-458
• /
• 2013

Temporal and spatial variations of water quality were investigated in the Cheonsu Bay of Yellow Sea, Korea from 2010 to 2011. Water samples were collected at 16 stations and physicochemical parameters were analyzed including water temperature, salinity, suspended solids (SS), chemical oxygen demand (COD), dissolved oxygen (DO), Chlorophyll a and nutrients. Spatial distribution patterns of all survey items were not clear among stations but the bimonthly variations were distinct except the bottom water of the suspended solids. The trend analysis by principal component analysis (PCA) during 2 years revealed the significant variations in water quality in the study area. Spatial water qualities were discriminated into 3 clusters by PCA; station cluster in the surface water 1, 2~11, and 12~16, the bottom water 1, 2~7, and 8~16. Annual bimonthly water qualities were clearly discriminated into 3 clusters by PCA. But tend of cluster in the surface and bottom water was difference, period most of the research was low in nutrient. Ecology-based water quality criteria was a good level of grade II. Bimonthly results are shown as III grade(normal) at June and August, II grade(good) at October and December and I grade for February and April. Water quality was showed by the input of fresh water same as those of Kyoungin coastal area, Asan coastal area, Gunsan coastal and Mokpo coastal area in the Cheonsu.

• Proceedings of the Plant Resources Society of Korea Conference
• /
• 2010.05a
• /
• pp.18-18
• /
• 2010

Glucosinolates(GSLs) are major secondary products($\beta$-thioglucoside N-hydroxysulfates) containing sulfates that are found mainly in Cruciferae family such as cabbage, broccoli, radish, turnip, kale, mustard and rapeseed as the important crops in agronomic and economic aspects. Especially, isothiocyanates(ITCs) have been shown to inhibit carcinogenesis with regulation cancer cell development followed by regulating target enzymes, inducing cell cycle arrest and apoptosis against various human cancer cell lines. Total desulfo(DS)-GSLs of nine type in 210 Korean leaf mustard accessions were isolated and confirmed using electrospray ionization mass spectrometric(ESI-MS/MS) method and DS-GSLs powerful library maked from several reference data. The isolated individual DS-GSLs were identified by removed $\beta$-D-glucopyranose residue ($C_6H_{11}O_5$, MW 163) from MS/MS data. Among them sinigrin(41.7%), glucoiberverin(21.7%) and gluconasturtiin(12.6%) were major components, Especially, despite gluconapin was minor component, accession K046197-1 and K046197-2 showed higher content of 4.11 and 3.31 mg/g(DW), respectively. The total GSLs contents in 210 accessions were ranged from 5.3 to 23.2 mg/g(DW) with a mean value of 13.0 mg/g(DW). As a result of principal component analysis(PCA), the individual GSLs loading plots were composed of three groups and components belonged to each group showed correlationship in quantitative pattern.

• Conservation Science in Museum
• /
• v.2
• /
• pp.69-76
• /
• 2000

This study, a part of scientific analysis of the 23 Neolithic red earthenwares excavated from southeastern coast and islands. We analyzed earthenware through statistical method and hardness analysis of clay, following the analysis of clay composition, its firing temperature, pigment composition and its thickness. Clay composition was analyzed by using ICP-AES and XRF, and then Principal Component Analysis (one of multivariate methods) was used for classification. As a result, clay of the earthenware was classified into 3 groups(Bumbang, Youkjido and other sites). In addition, hardness analysis of pigment and clay based on the depth of earthenware surface showed that pigment layer containing lots of Fe₂O₃ had higher hardness than clay part, which can be interpreted that Fe₂O₃ contributes to raising hardness in case of high temperature firing. The fact that pigment hardness is higher than that of clay part implies that pigment was applied before firing.

• Natural Product Sciences
• /
• v.17 no.2
• /
• pp.147-159
• /
• 2011

A high-performance liquid chromatography-electrospray ionization-tandem mass spectrometric method was developed to determine simultaneously eight marker constituents of Forsythiae fructus, and subsequently applied it to classify its two botanical origins. The marker compounds of Forsythia suspensa were phillyrin, pinoresinol, phillygenin, lariciresinol and forsythiaside; those of F.viridissima were arctiin, arctigenin and matairesinol. Separation of the eight analytes was achieved on a phenyl-hexyl column (150${\times}$2.0 mm i.d., 3 ${\mu}M$) using gradient elution with the mobile phase: (A) 10% acetonitrile in 0.5% acetic acid, (B) 40% aqueous acetonitrile. A few fragment ions specific to the types of lignans, among the product ions generated by collisonally induced dissociation (CID) of molecular ion clusters, such as [M-H]$^-$ or [M+OAc]$^-$ were used not only for fingerprinting analysis but for the quantification of each epimer by using multiple-reaction monitoring mode. It was shown good linearity ($r^2{\geq}$ 0.9998) over the wide range of all analytes; intra- and inter-day precisions (RSD, %) were within 9.14% and the accuracy ranged from 84.3 to 115.1%. The analytical results of 40 drug samples, combined with multivariate statistical analyses - principal component analysis (PCA) and hierarchical cluster analysis (HCA) - clearly demonstrated the classification of the test samples according to their botanical origins. This method would provide a practical strategy for assessing the authenticity or quality of the herbal drug.

• Bulletin of the Korean Chemical Society
• /
• v.30 no.1
• /
• pp.137-144
• /
• 2009

In this study, quantitative and pattern recognition analyses for the quality evaluation of Herba Epimedii using HPLC was developed. For quantitative analysis, five major bioactive constituents, hyperin, epimedin A, epimedin B, epimedin C, and icariin were determined. Analysis was carried out on Capcell pak $C_{18}$ column ($250{\time}4.6$ mm, 5 ${\mu}m$) with a mobile phase of mixture of acetonitrile and 0.1% formic acid, using UV detection at 270 nm. The linear behavior was observed over the investigated concentration range (2-50 ${\mu}g/mL;\;r_2\;>$ 0.99) for all analytes. The intraand inter-day precisions were lower than 4.3% (as a relative standard deviation, RSD) and accuracies between 95.1% and 104.4%. The HPLC analytical method for pattern recognition analysis was validated by repeated analysis of one reference sample. The RSD of intra- and inter-day variation of relative retention time (RRT) and relative peak area (RPA) of the 12 selected common peaks were below 0.8% and 4.7%, respectively. The developed methods were applied to analysis of twenty Herba Epimedii extract samples. Contents of hyperin, epimedin A, epimedin B, epimedin C, and icariin were calculated to be 0$\sim$0.79, 0.69$\sim$1.91, 0.93$\sim$9.58, 0.65$\sim$3.05, and 2.43$\sim$11.8 mg/g dried plant. Principal component analysis (PCA) showed that most samples were clustered together with the reference samples but several apart from the main cluster in the PC score plot, indicating differences in overall chemical composition between two clusters. The present study suggests that quantitative determination of marker compounds combined with pattern-recognition method can provide a comprehensive approach for the quality assessment of herbal medicines.