• Journal of information and communication convergence engineering
• /
• v.21 no.1
• /
• pp.24-31
• /
• 2023

The solid-state drive (SSD) possesses higher input and output speeds, more resistance to physical shock, and lower latency compared with regular hard disks; hence, it is an increasingly popular storage device. However, tiny components on an internal printed circuit board (PCB) hinder the manual detection of malfunctioning components. With the rapid development of artificial intelligence technologies, automatic detection of components through convolutional neural networks (CNN) can provide a sound solution for this area. This study proposes applying the YOLOv5 model to SSD PCB component detection, which is the first step in detecting defective components. It achieves pioneering state-of-the-art results on the SSD PCB dataset. Contrast experiments are conducted with YOLOX, a neck-and-neck model with YOLOv5; evidently, YOLOv5 obtains an mAP@0.5 of 99.0%, essentially outperforming YOLOX. These experiments prove that the YOLOv5 model is effective for tiny object detection and can be used to study the second step of detecting defective components in the future.

• International Journal of Computer Science & Network Security
• /
• v.24 no.5
• /
• pp.129-134
• /
• 2024

Object Detection is an emerging technology in the field of Computer Vision and Image Processing that deals with detecting objects of a particular class in digital images. It has considered being one of the complicated and challenging tasks in computer vision. Earlier several machine learning-based approaches like SIFT (Scale-invariant feature transform) and HOG (Histogram of oriented gradients) are widely used to classify objects in an image. These approaches use the Support vector machine for classification. The biggest challenges with these approaches are that they are computationally intensive for use in real-time applications, and these methods do not work well with massive datasets. To overcome these challenges, we implemented a Deep Learning based approach Convolutional Neural Network (CNN) in this paper. The Proposed approach provides accurate results in detecting objects in an image by the area of object highlighted in a Bounding Box along with its accuracy.

• Journal of Sensor Science and Technology
• /
• v.28 no.2
• /
• pp.94-100
• /
• 2019

Drivable area detection is a major task in advanced driver assistance systems. For drivable area detection, several studies have proposed vision-sensor-based approaches. However, conventional drivable area detection methods that use vision sensors are not suitable for environments with changes in road elevation. In addition, if the boundary between the road and vegetation is not clear, judging a vegetation area as a drivable area becomes a problem. Therefore, this study proposes an accurate method of detecting drivable areas in environments in which road elevations change and vegetation exists. Experimental results show that when compared to the conventional method, the proposed method improves the average accuracy and recall of drivable area detection on the KITTI vision benchmark suite by 3.42%p and 8.37%p, respectively. In addition, when the proposed vegetation area removal method is applied, the average accuracy and recall are further improved by 6.43%p and 9.68%p, respectively.

• Korean Journal of Radiology
• /
• v.19 no.6
• /
• pp.1007-1020
• /
• 2018

Objective: To compare the diagnostic performance of cardiovascular magnetic resonance (CMR) myocardial perfusion at 1.5- and 3-tesla (T) for detecting significant coronary artery disease (CAD), with invasive coronary angiography (ICA) as the reference method. Materials and Methods: We prospectively enrolled 281 patients (age $62.4{\pm}8.3$ years, 193 men) with suspected or known CAD who had undergone 1.5T or 3T CMR and ICA. Two independent radiologists interpreted perfusion defects. With ICA as the reference standard, the diagnostic performance of 1.5T and 3T CMR for identifying significant CAD (${\geq}50%$ diameter reduction of the left main and ${\geq}70%$ diameter reduction of other epicardial arteries) was determined. Results: No differences were observed in baseline characteristics or prevalence of CAD and old myocardial infarction (MI) using 1.5T (n = 135) or 3T (n = 146) systems. Sensitivity, specificity, positive and negative predictive values, and area under the receiver operating characteristic curve (AUC) for detecting significant CAD were similar between the 1.5T (84%, 64%, 74%, 76%, and 0.75 per patient and 68%, 83%, 66%, 84%, and 0.76 per vessel) and 3T (80%, 71%, 71%, 80%, and 0.76 per patient and 75%, 86%, 64%, 91%, and 0.81 per vessel) systems. In patients with multi-vessel CAD without old MI, the sensitivity, specificity, and AUC with 3T were greater than those with 1.5T on a per-vessel basis (71% vs. 36%, 92% vs. 69%, and 0.82 vs. 0.53, respectively). Conclusion: 3T CMR has similar diagnostic performance to 1.5T CMR in detecting significant CAD, except for higher diagnostic performance in patients with multi-vessel CAD without old MI.

• Journal of Food Hygiene and Safety
• /
• v.38 no.5
• /
• pp.279-286
• /
• 2023

Rapid and accurate detection of pathogenic bacteria is crucial for various applications, including public health and food safety. However, existing bacteria detection techniques have several drawbacks as they are inconvenient and require time-consuming procedures and complex machinery. Recently, the precision and versatility of CRISPR/Cas system has been leveraged to design biosensors that offer a more efficient and accurate approach to bacterial detection compared to the existing techniques. Significant research has been focused on developing biosensors based on the CRISPR/Cas system which has shown promise in efficiently detecting pathogenic bacteria or virus. In this review, we present a biosensor based on the CRISPR/Cas system that has been specifically developed to overcome these limitations and detect different pathogenic bacteria effectively including Vibrio parahaemolyticus, Salmonella, E. coli O157:H7, and Listeria monocytogenes. This biosensor takes advantage of the CRISPR/Cas system's precision and versatility for more efficiently accurately detecting bacteria compared to the previous techniques. The biosensor has potential to enhance public health and ensure food safety as the biosensor's design can revolutionize method of detecting pathogenic bacteria. It provides a rapid and reliable method for identifying harmful bacteria and it can aid in early intervention and preventive measures, mitigating the risk of bacterial outbreaks and their associated consequences. Further research and development in this area will lead to development of even more advanced biosensors capable of detecting an even broader range of bacterial pathogens, thereby significantly benefiting various industries and helping in safeguard human health

• Proceedings of the Korean Geotechical Society Conference
• /
• 2004.03b
• /
• pp.904-909
• /
• 2004

The study area adjoins with Yangsan fault in Sangbuk-myeon, Samsam-ri, Kyongsang-namdo and consist of the natural steep slope. After drawing data layer which have altitude by using digital topography data, it is converted to lattice DEM of $10m{\times}10m$ size. From this, gradient map of unit lattice, slant direction map and shadow relif map are made. Using flow apportioning algorithm, upper slope contributing area and wetness index by established lattice can be calculated. Area that have high wetness index shows lineament structure of northwest-southeast direction, and this agrees with shear fracture system. The result of electricity specific resistance survey in the study area shows that area of high wetness index has low electricity specific resistance anomaly. That is, wetness index conforms with distribution of fractured zone that accompanied chemical weathering of rock. Therefore, wetness index can be used as the method of detecting fractured zones and judging the stability of the area.

• Journal of the Ergonomics Society of Korea
• /
• v.34 no.5
• /
• pp.437-445
• /
• 2015

Objective: The aims of this study are to investigate the trends of intellectual property in order to identify the ergonomic approaches on musculoskeletal disorders, harmful factors of musculoskeletal disorders, and to find the potential applicability of motion capture technology. Background: Ergonomic posture assessment tools often showed interrater variance, though the usage is easy and practical in industrial fields. Moreover new technologies such as motion capture showed the potential applicability in posture assessment. So ergonomists and practitioners became interested in the intellectual properties on musculoskeletal disorder and motion capture technology. Method: Intellectual properties were collected with the combination of keywords such as ergonomic, musculoskeletal disorder, and motion capture using the KIPRIS (Korea Intellectual Property Rights Information Service). Collected intellectual properties were classified into ergonomic area and non-ergonomic area, except unexamined intellectual properties. This study investigated the trend of application of intellectual properties and the probability of using motion capture technology. Results: Few intellectual properties with ergonomic approach on musculoskeletal disorders were founded, despite many products for rehabilitation and sports. One hundred twenty five patents in 1105 patents on musculoskeletal disorders and 138 patents in 1908 patents on motion capture technology were classified into the patents that ergonomic approach can be applied. The patents related to ergonomics area are rapidly increasing after 2010, and there are good opportunities for ergonomists to apply the patents. Conclusion: This study found opportunities on novel methodology in detecting the harmful factors of musculoskeletal disorders, and that the motion capture technology is applicable in ergonomic posture assessment. Application: The results of this study can help ergonomists prepare the ergonomic patents, and can show the potential use of motion capture technology in detecting the harmful posture of musculoskeletal disorders.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.18 no.3
• /
• pp.315-322
• /
• 2000

DSM(Digital Surface Model) is a digital model which represents the surface elevation of a region. DSM is necessary for orthoimage generation, and frequently used in man-made object extraction from aerial photographs nowadays. Image matching technique enables automatic DSM generation. This proposed a image matching method which can be applied to automatic generation of DSM for Built-up Area. The matching method proposed is to find conjugate points and conjugate lines from overlapping aerial images. In detecting conjugate points, the positional relation between possible conjugate point pair as well as correlation of pixel gray value is compared. In detecting conjugate lines, the color attribute of flank region of line, shape of line, positional relation between neighborhood points and lines, and the connection relation between lines are compared. The proposed matching method is assumed to be useful for DSM generation including Built-up Area.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.5
• /
• pp.34-41
• /
• 2021

Sidewalks are facilities used for the safe and comfortable passage of pedestrians and are paved with blocks of various materials. Currently, Korea does not have a quantitative survey method for the pavement condition of sidewalks, so it is necessary to develop an efficient survey method. Drones are being used as an efficient survey tool in various fields, but there are limited studies in which sidewalks have been investigated. This study investigates the possibility of detection by limiting the missing sidewalk blocks using a drone. This study is an initial study on the development of a method for detecting damage in sidewalk blocks. For this, sidewalk blocks were artificially removed to simulate a dropout situation, and images were acquired with 0.7-cm resolution using a drone. As a characteristic of the point cloud data acquired through image pre-processing, there was high variance of the elevation of the points in the missing area of the sidewalk block. Using these characteristics, an experiment was conducted to detect the missing parts of the sidewalk block by applying four thresholds to the variance of the elevation of points included in the grid corresponding to the sidewalk area. As a result, the detection accuracy was shown with a positive detection ratio of 70-80%, omission errors of 20-30%, and commission errors lower than 2%. It is judged that the possibility of detecting missing sidewalk blocks is high. This study focused on detecting a simulated missing sidewalk block in a limited environment. Therefore, it is expected that an efficient and quantitative method of detecting damaged sidewalk blocks can be developed in the future through additional research with considerations of the actual environment.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.567-569
• /
• 2006

This study proposes the method to make a map of moving environment area and passed area by sensor using for recognizing environment or avoiding obstacle. We win develop an efficient algorithm to use sensors and get the data by the user friendly system. Through this system, we win study the way to know of the driving environment of moving robot on the long distance point. To this, we use only one ultra-sonic sensor with a servo motor which rotates 180 degree and loads an ultra-sonic sensor. A sensor can measure 1m${\sim}$8m range and a servo motor can distinguish 15 degree by 12 divide of 180 degree. By this feature of operating system, the robot which has these sensor module detects around of area and moves another point. In this way, users gather the data of detecting distance and change the data to X-Y coordinates. And users derive a map from these accumulate data.