• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 한국정보통신학회 2021년도 추계학술대회
• /
• pp.15-17
• /
• 2021

This paper proposes a drone flight trajectory method for accurate air pollutant emission designation detection. In areas with many factories, such as industrial complexes, there are workplaces that illegally emit air pollutants in a situation where monitoring is neglected. In the past, studies have been actively conducted to measure air pollutants in these areas using drones. The measurement method using a drone uses a method of detecting pollution by stopping around the chimney of a factory, but it has a problem in that the detection of air pollutants is inaccurate depending on environmental factors such as air pressure and wind. Therefore, this paper proposes a drone flight trajectory method for accurate air pollutant emission designation detection. This paper devises a screw orbit flight method in which a drone flies upward while rotating the chimney, and the total area of the chimney is detected and measured considering environmental factors. In the experiment, our proposal shows a higher performance than the existing method.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• 제48권1호
• /
• pp.41-49
• /
• 2022

Objectives: We compared changes in fractal dimension (FD) and grayscale value (GSV) of peri-implant alveolar bone on digital panoramic radiography (DPR) and cone-beam computed tomography (CBCT) immediately after implant surgery and 12 months postoperative. Materials and Methods: In this retrospective study, 16 patients who received posterior mandibular area dental implants with CBCT scans taken about 2 weeks after implantation and one year after implantation were analyzed. A region of interest was selected for each patient. FDs and GSVs were evaluated immediately after implant surgery and at 12-month follow-up to examine the functional loading of the implants. Results: There were no significant differences between DPR and CBCT measurements of FD values (P>0.05). No significant differences were observed between FD values and GSVs calculated after implant surgery and at the 12-month follow-up (P>0.05). GSVs were not correlated with FD values (P>0.05). Conclusion: The DPR and reconstructed panoramic CBCT images exhibit similar image quality for the assessment of FD. There were no changes in FD values or GSVs of the peri-implant trabecular bone structure at the 12-month postoperative evaluation of the functional loading of the implant in comparison to values immediately after implantation. GSVs representing bone mass do not align with FD values that predict bone microstructural parameters. Therefore, GSVs and FDs should be considered different parameters for assessing bone quality.

• Korean Journal of Remote Sensing
• /
• 제40권1호
• /
• pp.57-69
• /
• 2024

Crop classification plays a vitalrole in monitoring agricultural landscapes and enhancing food production. In this study, we explore the effectiveness of Long Short-Term Memory (LSTM) models for crop classification, focusing on distinguishing between apple and rice crops. The aim wasto overcome the challenges associatedwith finding phenology-based classification thresholds by utilizing LSTM to capture the entire Normalized Difference Vegetation Index (NDVI)trend. Our methodology involvestraining the LSTM model using a reference site and applying it to three separate three test sites. Firstly, we generated 25 NDVI imagesfrom the Sentinel-2A data. Aftersegmenting study areas, we calculated the mean NDVI values for each segment. For the reference area, employed a training approach utilizing the NDVI trend line. This trend line served as the basis for training our crop classification model. Following the training phase, we applied the trained model to three separate test sites. The results demonstrated a high overall accuracy of 0.92 and a kappa coefficient of 0.85 for the reference site. The overall accuracies for the test sites were also favorable, ranging from 0.88 to 0.92, indicating successful classification outcomes. We also found that certain phenological metrics can be less effective in crop classification therefore limitations of relying solely on phenological map thresholds and emphasizes the challenges in detecting phenology in real-time, particularly in the early stages of crops. Our study demonstrates the potential of LSTM models in crop classification tasks, showcasing their ability to capture temporal dependencies and analyze timeseriesremote sensing data.While limitations exist in capturing specific phenological events, the integration of alternative approaches holds promise for enhancing classification accuracy. By leveraging advanced techniques and considering the specific challenges of agricultural landscapes, we can continue to refine crop classification models and support agricultural management practices.

• Nuclear Engineering and Technology
• /
• 제55권10호
• /
• pp.3648-3658
• /
• 2023

The early localization of a fuel subassembly with a failed (wet rupture) fuel pin is very important in reactors to limit the associated radiological and operational consequences. This requires a fast and reliable system for failure detection and their localization in the core. In the Prototype Fast Breeder Reactor, the system specially designed for this purpose is Failed Fuel Location Modules (FFLM) housed in the control plug region. It identifies a failed sub-assembly by detecting the presence of delayed neutrons in the sodium from a failed sub-assembly. During the commissioning phase of PFBR, it is mandatory to demonstrate the FFLM effectiveness. The paper highlights the engineering and physics design aspects of FFLM and the integrated simulation towards its function demonstration with a source assembly containing a perforated metallic fuel pin. This test pin mimics a MOX pin of 1 cm² of geometrical defect area. At 10% power and 20% sodium flow rate, the counts rate in the BCCs of FFLM system range from 75 cps to 145 cps depending upon the position of DN source assembly. The model developed for the counts simulation is applicable to both metal and MOX pins with proper values of k-factor and escape coefficient.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권12호
• /
• pp.3218-3241
• /
• 2023

Financial fraud undermines the sustainable development of financial markets. Financial statements can be regarded as the key source of information to obtain the operating conditions of listed companies. Current research focuses more on mining financial digital data instead of looking into text data. However, text data can reveal emotional information, which is an important basis for detecting financial fraud. The audit opinion of the financial statement is especially the fair opinion of a certified public accountant on the quality of enterprise financial reports. Therefore, this research was carried out by using the data features of 4,153 listed companies' financial annual reports and audits of text opinions in the past six years, and the paper puts forward a financial fraud detection model integrating audit opinions. First, the financial data index database and audit opinion text database were built. Second, digitized audit opinions with deep learning Bert model was employed. Finally, both the extracted audit numerical characteristics and the financial numerical indicators were used as the training data of the LightGBM model. What is worth paying attention to is that the imbalanced distribution of sample labels is also one of the focuses of financial fraud research. To solve this problem, data enhancement and Focal Loss feature learning functions were used in data processing and model training respectively. The experimental results show that compared with the conventional financial fraud detection model, the performance of the proposed model is improved greatly, with Area Under the Curve (AUC) and Accuracy reaching 81.42% and 78.15%, respectively.

• Journal of the Korea Society of Computer and Information
• /
• 제29권1호
• /
• pp.77-84
• /
• 2024

In this paper, we propose a method for accurately and rapidly detecting defects in wire harnesses by utilizing computer vision to calculate six crucial measurement values: the length of crimped terminals, the dimensions (width) of terminal ends, and the width of crimped sections (wire and core portions). We employ Harris corner detection to locate object positions from two types of data. Additionally, we generate reference points for extracting measurement values by utilizing features specific to each measurement area and exploiting the contrast in shading between the background and objects, thus reflecting the slope of each sample. Subsequently, we introduce a method using the Euclidean distance and correction coefficients to predict values, allowing for the prediction of measurements regardless of changes in the wire's position. We achieve high accuracy for each measurement type, 99.1%, 98.7%, 92.6%, 92.5%, 99.9%, and 99.7%, achieving outstanding overall average accuracy of 97% across all measurements. This inspection method not only addresses the limitations of conventional visual inspections but also yields excellent results with a small amount of data. Moreover, relying solely on image processing, it is expected to be more cost-effective and applicable with less data compared to deep learning methods.

• The Journal of the Acoustical Society of Korea
• /
• 제27권2호
• /
• pp.80-86
• /
• 2008

Breast ultrasound has many advantages over mammography but suffers from a shortcoming of being not suitable in detecting microcalcification. We studied on a method based on acoustic resonance and power Doppler to detect calcification of breast tissue using a typical 7.5 MHz linear probe used in breast ultrasound examination. We first constructed a breast tissue phantom made of gelatin and then observed calcified legions as external vibrations varied. Calcification injected to the breast tissue phantom being resonated different from the surrounding medium, and its acoustic resonance driven by external vibrations was visualized by differences for color brightness and area in ROI of power doppler. In low frequency regions, the acoustic resonance almost not appeared and showed a plateau in $300{\sim}600\;Hz$ and the color vanished as the frequency further increased.

• Journal of Advanced Navigation Technology
• /
• 제27권6호
• /
• pp.888-896
• /
• 2023

When an aircraft is landing, it is expected that the aircraft will follow a specified approach procedure and then land at the airport. However, depending on the airport situation, neighbouring aircraft or the instructions of the air traffic controller, there can be a deviation from the specified approach. Detecting aircraft approach patterns is necessary for traffic flow and flight safety, and this paper suggests clustering techniques to identify aircraft patterns in the approach segment. The Gaussian Mixture Model (GMM), one of the machine learning techniques, is used to cluster the trajectories of aircraft, and ADS-B data from aircraft landing at the Gimhae airport in 2019 are used. The aircraft trajectories are clustered on the plane, and a total of 86 approach trajectory patterns are extracted using the centroid value of each cluster. Considering the correlation between the approach procedure pattern and overshoots, the distribution of overshoots is calculated.

• Korean Journal of Radiology
• /
• 제21권7호
• /
• pp.891-899
• /
• 2020

Objective: To assess the diagnostic performance of a deep learning-based algorithm for automated detection of acute and chronic rib fractures on whole-body trauma CT. Materials and Methods: We retrospectively identified all whole-body trauma CT scans referred from the emergency department of our hospital from January to December 2018 (n = 511). Scans were categorized as positive (n = 159) or negative (n = 352) for rib fractures according to the clinically approved written CT reports, which served as the index test. The bone kernel series (1.5-mm slice thickness) served as an input for a detection prototype algorithm trained to detect both acute and chronic rib fractures based on a deep convolutional neural network. It had previously been trained on an independent sample from eight other institutions (n = 11455). Results: All CTs except one were successfully processed (510/511). The algorithm achieved a sensitivity of 87.4% and specificity of 91.5% on a per-examination level [per CT scan: rib fracture(s): yes/no]. There were 0.16 false-positives per examination (= 81/510). On a per-finding level, there were 587 true-positive findings (sensitivity: 65.7%) and 307 false-negatives. Furthermore, 97 true rib fractures were detected that were not mentioned in the written CT reports. A major factor associated with correct detection was displacement. Conclusion: We found good performance of a deep learning-based prototype algorithm detecting rib fractures on trauma CT on a per-examination level at a low rate of false-positives per case. A potential area for clinical application is its use as a screening tool to avoid false-negative radiology reports.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• 제24권3호
• /
• pp.163-169
• /
• 2024

Anti-drone (anti-drone) is at the center of the debate over the failure to shoot down a North Korean drone that invaded the metropolitan area at the end of 2022. Anti-drone is a means of detecting and restraining drone flights in unauthorized airspace. Anti-drone technology is a key defense system for drone technology that is essential in the current illegal situation of various drones. We must be alert in the war in Ukraine, where the role of drones has increased. Drone attacks, which are not easy to defend, may determine the victory or defeat of the war. Competition for anti-drone technology development in countries around the world will rise. When new anti-drone technology emerges, drones that go beyond it will be developed. This study presented the current status of anti-drone by analyzing the defense system of domestic drones.