• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.4D
• /
• pp.687-693
• /
• 2006

Image co-registration is the process of overlaying two images of the same scene, one of which represents a reference image, while the other is geometrically transformed to the one. In order to improve efficiency and effectiveness of the co-registration approach, the author proposed a pre-qualified area matching algorithm which is composed of feature extraction with canny operator and area matching algorithm with cross correlation coefficient. For refining matching points, outlier detection using studentized residual was used and iteratively removes outliers at the level of three standard deviation. Throughout the pre-qualification and the refining processes, the computation time was significantly improved and the registration accuracy is enhanced. A prototype of the proposed algorithm was implemented and the performance test of 3 Landsat images of Korea. showed: (1) average RMSE error of the approach was 0.435 pixel; (2) the average number of matching points was over 25,573; (3) the average processing time was 4.2 min per image with a regular workstation equipped with a 3 GHz Intel Pentium 4 CPU and 1 Gbytes Ram. The proposed approach achieved robustness, full automation, and time efficiency.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.85-92
• /
• 2023

In this paper, we propose a method for automatically generating story-based abbreviated summaries from long-form dramas and movies. From the shooting stage, the basic premise was to compose a frame with illusion of depth considering the golden division as well as focus on the object of interest to focus the viewer's attention in terms of content delivery. To consider how to extract the appropriate frames for this purpose, we utilized elemental techniques that have been utilized in previous work on scene and shot detection, as well as work on identifying focus-related blur. After converting the videos shared on YouTube to frame-by-frame, we divided them into a entire frame and three partial regions for feature extraction, and calculated the results of applying Laplacian operator and FFT to each region to choose the FFT with relative consistency and robustness. By comparing the calculated values for the entire frame with the calculated values for the three regions, the target frames were selected based on the condition that relatively sharp regions could be identified. Based on the selected results, the final frames were extracted by combining the results of an offline change point detection method to ensure the continuity of the frames within the shot, and an edit decision list was constructed to produce an abbreviated summary of 62.77% of the footage with F1-Score of 75.9%

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.2
• /
• pp.119-125
• /
• 2023

In this study, we photograph driving obstacle objects such as crosswalks, side spheres, manholes, braille blocks, partial ramps, temporary safety barriers, stairs, and inclined curb that hinder or cause inconvenience to the movement of the vulnerable using electric mobility aids. We develop an optimal AI model that classifies photographed objects and automatically recognizes them, and implement an algorithm that can efficiently determine obstacles in front of electric mobility aids. In order to enable object detection to be AI learning with high probability, the labeling form is labeled as a polygon form when building a dataset. It was developed using a Mask R-CNN model in Detectron2 framework that can detect objects labeled in the form of polygons. Image acquisition was conducted by dividing it into two groups: the general public and the transportation weak, and image information obtained in two areas of the test bed was secured. As for the parameter setting of the Mask R-CNN learning result, it was confirmed that the model learned with IMAGES_PER_BATCH: 2, BASE_LEARNING_RATE 0.001, MAX_ITERATION: 10,000 showed the highest performance at 68.532, so that the user can quickly and accurately recognize driving risks and obstacles.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.25 no.6
• /
• pp.555-567
• /
• 2023

The cracks in the tunnel are currently determined through visual inspections conducted by inspectors based on images acquired using tunnel imaging acquisition systems. This labor-intensive approach, relying on inspectors, has inherent limitations as it is subject to their subjective judgments. Recently research efforts have actively explored the use of deep learning to automatically detect tunnel cracks. However, most studies utilize public datasets or lack sufficient objectivity in the analysis process, making it challenging to apply them effectively in practical operations. In this study, we selected test datasets consisting of images in the same format as those obtained from the actual inspection system to perform an objective evaluation of deep learning models. Additionally, we introduced ensemble techniques to complement the strengths and weaknesses of the deep learning models, thereby improving the accuracy of crack detection. As a result, we achieved high recall rates of 80%, 88%, and 89% for cracks with sizes of 0.2 mm, 0.3 mm, and 0.5 mm, respectively, in the test images. In addition, the crack detection result of deep learning included numerous cracks that the inspector could not find. if cracks are detected with sufficient accuracy in a more objective evaluation by selecting images from other tunnels that were not used in this study, it is judged that deep learning will be able to be introduced to facility safety inspection.

• Korean journal of applied entomology
• /
• v.29 no.3
• /
• pp.209-215
• /
• 1990

Some preliminary studies were conducted to find out whether the levee-burning could justifiable for the suppression of insect pests, particularly the smaller brown planthopper (Laodelphax striatellus F.). Density surveys on pests and their enemies (mostly spiders) were carried out upto the mid May at an experimental paddy field located in Suwon after of it's levee $(72\times1m)$ was burned on Feb. 20, 1987. Results were discussed in relation to density recovering of both pests and their possible enemies (spiders) and summarized as below. Not a single individual of any pest or enemy was found from the levee upto sometime after the levee-burning. Grasses started to grow more vigorously in burned ares than in unburned upto about 60 days after the burning. And densities of both pest and enemies grew higher in burned areas than in unburned from about 75 days after the burning (in Early may). It is suspected that all individuals of pests and enemies fond from the burned areas could have immigrated from the surrounding areas. If levee-burning was carried out in much wider areas, much longer time would be needed to recover the density of both pests and enemies to the center region of the burning. Wingless spiders would require even longer time than winged pest species to re-establish in the center region of the widely burned field. Pirata subpiraticus, the most abundant spider species in Korean paddy fields, starts to move about and searches for food at above $9^{\circ}C$ which is somewhat lower than the critical temperature for the pest species. Thus P. subpiraticus would require more food than other pest species early in the spring, and therefore, it would have lower probability to survive than pest species particularly in burned areas. Experiments for pest suppression with levee-burning would better be carried on in much wider areas, and its justification seems to be discussed after man other disciplines related to both pests and their natural enemies were throughly studied together with their density surveys. However, according to the present point of vie, the opinion that levee-burning is helpful for controlling pests which over winter on levee areas could not be justifiable.

• Korean Journal of Clinical Laboratory Science
• /
• v.39 no.2
• /
• pp.113-121
• /
• 2007

Hearing loss is a common congenital disorder that is frequently associated with mutations in the Cx26 gene (GJB2). Recently, the mutation analysis of GJB2 has been used in a newborn screening test for the detection of hearing impairment. Population-based studies should be performed before the application of genetic testing for the identification of deaf newborns. In this study, 8 positions of GJB2 mutations-including 35delG, 167delT, 235delC, V27I, V37I, M34T, E114G, and I203T-were analyzed using PCR-direct sequencing in a total of 437 healthy Korean neonates. DNAs from dried blood spots were extracted using a commercial DNA extraction kit. The PCR-amplified products (783 bps) of the GJB2 gene were detected using 2% agarose gel electrophoresis and subjected to direct sequencing. The sequences were compared with those in the GenBank database by using the BLAST program. In this study, 5 GJB2 mutations -including V27I (79G>A), V37I (109G>A), E114G (341A>G), I203T (608T>C), and 235delC- were found. Of the 437 neonate samples, 301 subjects showed GJB2 mutations (68.9%, 301/437). The V27I mutation was found in 271 subjects and was the most frequent (62.0%, 271/437). The E114G, I203T and V37I mutations were shown in 146, 17 and 14 subjects, respectively. The 235delC mutation was found in 1 subject. The E114G mutation was frequently accompanied by the V27I mutation. V27I/E114G (97.2%, 143/147) was the most common double mutation and 3 subjects had the double mutation V27I/I203T. A triple mutation, V27I/E114G/I203T, was found in 1 subject. In conclusion, PCR-direct sequencing is a convenient tool for the rapid detection of GJB2 mutations and this data might provide information for the genetic counseling of the GJB2 gene.

• Korean Journal of Clinical Laboratory Science
• /
• v.38 no.3
• /
• pp.196-202
• /
• 2006

Treatment of hepatitis B virus (HBV) with lamivudine is effective in suppressing virus replication and results in reduced inflammatory activity. However the most troublesome problem of lamivudine treatment is the emergence of lamivudine-resistant strains with amino acid substitution in the YMDD motif of DNA polymerase gene during the treatment. The aim of this study was to determine the mutation of YMDD motif (codon 552) and codon 528 in chronic HBV patients with lamivudine therapy using PCR-direct sequencing and to investigate the relationship between lamivudine mediated HBV mutation and HBeAg. HBV DNA was extracted from serum samples of HBV patients and amplified by nested PCR with two sets of primer pairs selected in HBV DNA polymerase gene. Amplified PCR product was analyzed by 2% agarose gel electrophoresis and direct sequencing. HBV mutation was detected in 124 out of 207 samples (60%). Single mutation was 50.8% for M552I, 43.5% for M552V, 5.7% for M552I/V and the L528M mutation was 67.0%. Double mutation was 43.6% for M552V/L528M, 33.1% for M552I/L528(wild type), 17.7% for M552I/L528M and 5.6% for M552I/V/L528M. Serine mutation at YMDD motif (M552S) was not found and the L528M mutation frequently accompanied M552V type. In this study, the typical difference of frequencies for HBV mutation depending on HBeAg was not found. Moreover, the PCR-direct sequencing method used in this study might be a powerful tool for the mutation study in clinical reference laboratories with high volume.

• Journal of radiological science and technology
• /
• v.30 no.1
• /
• pp.13-23
• /
• 2007

In recent years, mammography system is changed rapidly from conventional screen-film system to digital system for application to screening and diagnosis. Digital mammography system provides several advantages over screen-film mammography system. According to the information provided by the manufacturer, digital mammography system offers radiation dose reduction in comparison with screen-film mammography system, because of digital detector, particularly direct digital detector has higher x-ray absorption efficiency than screen-film combination or imaging plate(IP). We measured average glandular doses(AGD) in screen-film mammography(SFM) system with slow screen-film combination, computed mammography(CM) system, indirect digital mammography(IDM) system and direct digital mammography(DDM) system using brest tissue-equivalent phantom(glandularity 30%, 50% and 70%). The results were shown as follows : AGD values for DDM system were highest than those for other systems. Although automatic exposure control(AEC) mode was selected, the curve of the AGD values against thickness or glandularity increased significantly for the SFM system with the uniform target/filter(Mo/Mo) combination. Therefore, the AGD values for the high energy examinations were highest in the SFM system, and those for the low energy examinations were highest in the DDM system. But the curve of the AGD values against thickness and glandularity increased gently for CM system with the automatic selection of the target/filter combination (from Mo/Mo to Mo/Rh or from Mo/Rh to Rh/Rh), and the AGD values were lowest. Consequently, the parameters in mammography for each exposure besides detection efficiency play an important role in oder to estimate a patient radiation dose.

• Journal of Korean Society of Forest Science
• /
• v.84 no.4
• /
• pp.415-423
• /
• 1995

This study was conducted to analyze the chemical composition, nutritional contents, and saponin in the xylem sap of Acer mono Max. From Feb 25 to Mar 4, 1994, spring sap was collected by making holes with 1.7cm diameter on the trunk of trees in Mt. Baekwoon and Mt. Jiri. Sugars were quantified by HPIC, amino acids by amino acid analyzer, saponins by TLC, HPLC, and $^1H$ & $^{13}C$ NMR. Major component of the sap was sucrose at a concentration range of 0.68 to 2.01%. Following minor components were found: glucose at 0.03-0.11%, and fructose at 0.01-0.03% as sugars, lipid at 0.03%, threonine at 0.152%, lysine at 0.038%, arginine at 0.068% as amino acids, ash at 0.1%, Ca at 175ppm, Fe at 2ppm, P at 19ppm, K at 16ppm, Na at 31ppm as minerals, vitamin $B_1$ at 0.6ppm, vit. $B_2$ at 0.1ppm, and vit. C at 19ppm. A trace of phenolic compounds was found by TLC, while saponin commonly reported in high-quality ginseng roots was not found in maple sap. It is concluded that sap of Acer mono contains a good variety of natural compounds such as sugars, amino acids, Ca, Fe, and vitamins to serve as an excellent source of very natural and health-promoting drink.

• Korean Journal of Remote Sensing
• /
• v.31 no.6
• /
• pp.501-512
• /
• 2015

Landfast sea ice (LFI) in Terra Nova Bay, East Antarctica where the Jangbogo Antarctic Research Station is located, has significant influences on marine ecosystem and the sailing of an icebreaker. Therefore, it is essential to analyze the spatio-temporal variation of the LFI in Terra Nova Bay. In this study, we chose interferometric pairs with the temporal baseline from 1 to 9 days out of a total of 62 COSMO-SkyMed synthetic aperture radar (SAR) images over Terra Nova Bay obtained from December 2010 to January 2012, and then constructed the coherence image of each pair. The LFI showed coherence values higher than 0.3 even in the interferometric SAR (InSAR) pairs of up to 9-days of temporal baseline. This was because the LFI was fixed at coastline and thus showed low temporal phase decorrelation. Based on the characteristics of the coherence on LFI, We defined the areas of LFI that show spatially homogeneous coherence values higher than 0.5. Pack ice (PI) and open water showed low coherence values due to large temporal phase decorreation caused by current and wind. Distinguishing PI from open water in the coherence images was difficult due to their similarly low coherence values. PI was identified in SAR amplitude images by investigating cracks on the ice. The extents of the LFI and PI were estimated from the coherence and SAR amplitude images and their temporal variations were analyzed. The extent of the LFI increased from March to July (maximum extent of $170.7km^2$) and decreased from October. The extent of the PI increased from February to May and decreased from May to July when the LFI increases dramatically. The extent of the LFI and air temperature showed an inverse correlation with a time lag of about 2 months, i.e., the extent of the LFI decreases after 2 months of the increase in the air temperature. Meanwhile the correlation between wind speed and the extent of the LFI was very low. This represents that the extent of LFI in Terra Nova Bay are influenced more by the air temperature than wind speed.