• Structural Engineering and Mechanics
• /
• v.91 no.5
• /
• pp.443-457
• /
• 2024

At present, the traditional concrete surface inspection methods based on artificial vision have the problems of high cost and insecurity, while the computer vision methods rely on artificial selection features in the case of sensitive environmental changes and difficult promotion. In order to solve these problems, this paper introduces deep learning technology in the field of computer vision to achieve automatic feature extraction of structural damage, with excellent detection speed and strong generalization ability. The main contents of this study are as follows: (1) A method based on DeepLabV3+ convolutional neural network model is proposed for surface detection of post-earthquake structural damage, including surface damage such as concrete cracks, spaling and exposed steel bars. The key semantic information is extracted by different backbone networks, and the data sets containing various surface damage are trained, tested and evaluated. The intersection ratios of 54.4%, 44.2%, and 89.9% in the test set demonstrate the network's capability to accurately identify different types of structural surface damages in pixel-level segmentation, highlighting its effectiveness in varied testing scenarios. (2) A semantic segmentation model based on DeepLabV3+ convolutional neural network is proposed for the detection and evaluation of post-earthquake structural components. Using a dataset that includes building structural components and their damage degrees for training, testing, and evaluation, semantic segmentation detection accuracies were recorded at 98.5% and 56.9%. To provide a comprehensive assessment that considers both false positives and false negatives, the Mean Intersection over Union (Mean IoU) was employed as the primary evaluation metric. This choice ensures that the network's performance in detecting and evaluating pixel-level damage in post-earthquake structural components is evaluated uniformly across all experiments. By incorporating deep learning technology, this study not only offers an innovative solution for accurately identifying post-earthquake damage in civil engineering structures but also contributes significantly to empirical research in automated detection and evaluation within the field of structural health monitoring.

• Economic and Environmental Geology
• /
• v.45 no.2
• /
• pp.169-180
• /
• 2012

The ESBs (Environmental Specimen Banks) have monitoring functions for the contemporary environmental qualities and also offer the future generation tangible information on the past environment by preserving the specimens. This entails the sampling of the representative specimen for each distinctive ecosystem, which is performed under a strict and stipulated procedure and a condition that does not allow any change in the component so that a retrospective analysis can be readily done even in the distant future. It has been more than 30 years that some developing countries started collecting a broad spectrum of specimens to vindicate the effectiveness of an environmental policy and to monitor the long-term variations of background concentrations of environmental pollutants. Though being late, the National Institute of Environmental Research (NIER), Korea, has successfully launched the National Environmental Specimen Bank (NESB) in 2009 equipped with its state-of-the-art automated cryogenic tanks. Since then, the researchers at the NESB have been doing their best to excel the existing ESBs around the world by learning and improving the expertise. To do so, they conducted a pilot study for developing and testing their own Standard Operating Procedure (SOP) based on the analysis of the examples of the other ESBs. The problems from the pilot study had been reviewed to improve the SOP to meet the requirements for an ESB, that is to say, preserving representative environmental specimens in cryogenic condition and enhancing the analytical method. Furthermore, they also need to prepare themselves to address the future challenges by providing some additional functions, which makes it distinguishable from the other ESBs. If successful, this will be a step further to be recognized as a full-fledged member of the ESB society of the world.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.3104-3104
• /
• 2001

The research described here was undertaken with the aim of monitoring, optimizing and ultimately controlling the production of heterofermentative microbes used as starters in the salami industry. The use of starter cultures in the fermented meats industry is a well-established technique used to shorten and standardize the ripening process, and to improve and control the organoleptic quality of the final product. Starter cultures are obtained by the submerged cultivation of suitable microorganisms in stirred, and sometimes aerated, fermenters where monitoring of key physiological parameters such as the concentration of biomass, substrates and metabolites suffers from the general lack of real-time measurement techniques applicable to aseptic processes. In this respect, the results of the present work are relevant to all submerged fermentation processes. Previous work on the application of on-line NIR spectroscopy to the lactic acid fermentation (Dosi et al. - Monreal NIR1995) had successfully used a system based on a measuring cell included in a circulation loop external to the fermenter. The fluid handling and sterility problems inherent in an external circulation system prompted us to explore the use of an in-line system where the NIR probe is immersed in the culture and is thus exposed to the hydrodynamic conditions of the stirred and aerated fluid. Aeration was expected to be a potential source of problems in view of the possible interference of air bubbles with the measurement device. The experimental set-up was based on an in-situ sterilizable NIR probe connected to the instrument by means of an optical fiber bundle. Preliminary work was carried out to identify and control potential interferences with the measurement, in particular the varying hydrodynamic conditions prevailing at the probe tip. We were successful in defining the operating conditions of the fermenter and the geometrical parameters of the probe (flow path, positioning, etc.) were the NIR readings were reliable and reproducible. The system thus defined was then used to construct and validate calibration curves for tile concentration of biomass, carbon source and major metabolites of two different microorganisms used as salami starters. Real-time measurement of such parameters coupled with the direct interfacing of the NIR instrument with the PC-based measurement and control system of the fermenter enabled the development of automated strategies for the interactive optimization of the starter production process.

• Korean Journal of Environmental Biology
• /
• v.34 no.2
• /
• pp.124-131
• /
• 2016

The toxic responses of flagellate Euglena agilis Carter to 8 heavy metals (Ag, Cd, Cr (VI), Cu, Hg, Ni, Pb, Zn) were measured using E. agilis system (E-Tox), an automated biomonitoring system. The E-Tox measures cell movement parameters, such as velocity, motility, and forms of the cells, as biological endpoints. $EC_{50}$ values from the E. agilis biomonitoring test were compared with the literature data from the tests with Daphnia magna, Vibrio fischeri and Euglena gracilis. Measurement of the E. agilis movement behavior and D. magna acute toxicity test were also conducted for the wastewater samples. E. agilis is less sensitive than D. magna but is comparable to or more sensitive than V. fischeri and E. gracilis for the heavy metals tested in this study. E. agilis shows prompt changes of these parameters for the toxic metal plating wastewater. Major advantages of the E-tox are automatic, easy to handle and fast ecotoxicity monitoring system compared to other biological test systems. These results imply that E. agilis biomonitoring test using E-Tox can be a putative ecotoxicity test as a good early warning tool for the monitoring of toxic wastewater.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.10
• /
• pp.63-68
• /
• 2017

Although automated metro subway systems have the advantage of operating a train without a train driver, it is difficult to detect an immediate fault condition and take countermeasures when an unusual situation occurs. Therefore, it is important to construct a maintenance information system (MIS) that detects the vehicle failure/status information in real time and maintains it efficiently in the depot of the railway's vehicles. This paper proposes a conceptual design method that realizes the interface between the train control system (TCS), the operation control center train control monitoring system (OCC-TCMS) console, and the MIS using wireless communication network in real-time. To transmit a large amount of information on 800,000 occurrences per day during operation, data was collected in a 56 byte data table using a data processing algorithm. This state information was classified into 4 hexadecimal codes and transmitted to the MIS by mapping the status and the fault information on the vehicle during the main line operation. Furthermore, the transmission and reception data were examined in real time between the TCS and MIS, and the implementation of the failure information screen was then displayed.

• Korean Journal of Remote Sensing
• /
• v.36 no.5_3
• /
• pp.1095-1107
• /
• 2020

Flood monitoring using satellite data has been constrained by obtaining satellite images for flood peak and accurately extracting flooded areas from satellite data. Deep learning is a promising method for satellite image classification, yet the potential of deep learning-based flooded area extraction using SAR data remained uncertain, which has advantages in obtaining data, comparing to optical satellite data. This research explores the performance of SegNet and U-Net on image segmentation by extracting flooded areas in the Khorat basin, Mekong river basin, and Cagayan river basin in Thailand, Laos, and the Philippines from Sentinel-1 A/B satellite data. Results show that Global Accuracy, Mean IoU, and Mean BF Score of SegNet are 0.9847, 0.6016, and 0.6467 respectively, whereas those of U-Net are 0.9937, 0.7022, 0.7125. Visual interpretation shows that the classification accuracy of U-Net is higher than SegNet, but overall processing time of SegNet is around three times faster than that of U-Net. It is anticipated that the results of this research could be used when developing deep learning-based flood monitoring models and presenting fully automated flooded area extraction models.

• Journal of Biosystems Engineering
• /
• v.28 no.6
• /
• pp.511-516
• /
• 2003

Computer vision technology has been utilized as one of the most powerful tools to automate various agricultural operations. Though it has demonstrated successful results in various applications, the current status of technology is still for behind the human's capability typically for the unstructured and variable task environment. In this paper, a man-machine interactive hybrid decision-making system which utilized a concept of tole-operation was proposed to overcome limitations of computer image processing and cognitive capability. Tasks of greenhouse watermelon cultivation such as pruning, watering, pesticide application, and harvest require identification of target object. Identifying water-melons including position data from the field image is very difficult because of the ambiguity among stems, leaves, shades. and fruits, especially when watermelon is covered partly by leaves or stems. Watermelon identification from the cultivation field image transmitted by wireless was selected to realize the proposed concept. The system was designed such that operator(farmer), computer, and machinery share their roles utilizing their maximum merits to accomplish given tasks successfully. And the developed system was composed of the image monitoring and task control module, wireless remote image acquisition and data transmission module, and man-machine interface module. Once task was selected from the task control and monitoring module, the analog signal of the color image of the field was captured and transmitted to the host computer using R.F. module by wireless. Operator communicated with computer through touch screen interface. And then a sequence of algorithms to identify the location and size of the watermelon was performed based on the local image processing. And the system showed practical and feasible way of automation for the volatile bio-production process.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.4
• /
• pp.979-991
• /
• 2016

Measuring nation images was a challenging task when employing offline surveys was the only option. It was not only prohibitively expensive, but too much time-consuming and therefore unfitted to this rapidly changing world. Although demands for monitoring real-time nation images were ever-increasing, an affordable and reliable solution to measure nation images has not been available up to this date. The researcher in this study developed a semi-automatic ontology construction algorithm, named "double-crossing double keyword collection (or DCDKC)" to measure nation images from Wikipedia in real-time. The ontology, WikiOnto, can be used to reflect dynamic image changes. In this study, an instance of WikiOnto was constructed by applying the algorithm to the big-three exporting countries in East Asia, Korea, Japan, and China. Then, the numbers of page views for words in the instance of WikiOnto were counted. A collection of the counting for each country was compared to each other to inspect the possibility to use for dynamic nation images. As for the conclusion, the result shows how the images of the three countries have changed for the period the study was performed. It confirms that DCDKC can very well be used for a real-time nation-image monitoring system.

• The Journal of Bigdata
• /
• v.8 no.2
• /
• pp.73-82
• /
• 2023

With 78% of current fisheries workers being elderly, there's a pressing need to address labor shortages. Consequently, active research on smart aquaculture technologies, centered on object detection and tracking algorithms, is underway. These technologies allow for fish size analysis and behavior pattern forecasting, facilitating the development of real-time monitoring and automated systems. Our study utilized video data from cameras outside aquaculture facilities and implemented fish detection and tracking algorithms. We aimed to tackle high maintenance costs due to underwater conditions and camera corrosion from ammonia and pH levels. We evaluated the performance of a real-time system using YOLOv7 for fish detection and the SORT algorithm for movement tracking. YOLOv7 results demonstrated a trade-off between Recall and Precision, minimizing false detections from lighting, water currents, and shadows. Effective tracking was ascertained through re-identification. This research holds promise for enhancing smart aquaculture's operational efficiency and improving fishery facility management.

• Research in Plant Disease
• /
• v.30 no.3
• /
• pp.256-267
• /
• 2024

We have developed FBcastS (Fire Blight Forecasting System), a cloud-based information system that leverages the K-Maryblyt forecasting model. The FBcastS provides an optimal timing for spraying antibiotics to prevent flower infection caused by Erwinia amylovora and forecasts the onset of disease symptoms to assist in scheduling field scouting activities. FBcastS comprises four discrete subsystems tailored to specific functionalities: meteorological data acquisition and processing, execution of the K-Maryblyt model, distribution of web-based information, and dissemination of spray timing notifications. The meteorological data acquisition subsystem gathers both observed and forecasted weather data from 1,583 sites across South Korea, including 761 apple or pear orchards where automated weather stations are installed for fire blight forecast. This subsystem also performs post-processing tasks such as quality control and data conversion. The model execution subsystem operates the K-Maryblyt model and stores its results in a database. The web-based service subsystem offers an array of internet-based services, including weather monitoring, mobile services for forecasting fire blight infection and symptoms, and nationwide fire blight monitoring. The final subsystem issues timely notifications of fire blight spray timing alert to growers based on forecasts from the K-Maryblyt model, blossom status, pesticide types, and field conditions, following guidelines set by the Rural Development Administration. FBcastS epitomizes a smart agriculture internet of things (IoT) by utilizing densely collected data with a spatial resolution of approximately 4.25 km to improve the accuracy of fire blight forecasts. The system's internet-based services ensure high accessibility and utility, making it a vital tool in data-driven smart agricultural practices.