• Journal of the Korean Society of Radiology
• /
• v.18 no.1
• /
• pp.21-26
• /
• 2024

The one of the famous deep learning models for object detection task is you only look once version 5 (YOLOv5) framework based on the one stage architecture. In addition, YOLOv5 model indicated high performance for accurate lesion detection using the bottleneck CSP layer and skip connection function. The purpose of this study was to evaluate the performance of YOLOv5 framework according to various hyperparameters in position emission tomogrpahy (PET) phantom images. The dataset was obtained from QIN PET segmentation challenge in 500 slices. We set the bounding box to generate ground truth dataset using labelImg software. The hyperparameters for network train were applied by changing optimization function (SDG, Adam, and AdamW), activation function (SiLU, LeakyRelu, Mish, and Hardwish), and YOLOv5 model size (nano, small, large, and xlarge). The intersection over union (IOU) method was used for performance evaluation. As a results, the condition of outstanding performance is to apply AdamW, Hardwish, and nano size for optimization function, activation function and model version, respectively. In conclusion, we confirmed the usefulness of YOLOv5 network for object detection performance in nuclear medicine images.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.7 no.1
• /
• pp.41-48
• /
• 2007

Latest advances in network sensor technology and state of the art of mobile robot, and artificial intelligence research can be employed to develop autonomous and distributed monitoring systems. In this study, as the preliminary step for developing a multi-purpose "Intelligent Space" platform to implement advanced technologies easily to realize smart services to human. We will give an explanation for the ISpace system architecture designed and implemented in this study and a short review of existing techniques, since there exist several recent thorough books and review paper on this paper. Instead we will focus on the main results with relevance to the DIND data fusion with CI of Intelligent Space. We will conclude by discussing some possible future extensions of ISpace. It is first dealt with the general principle of the navigation and guidance architecture, then the detailed functions tracking multiple objects, human detection and motion assessment, with the results from the simulations run.

• Journal of Korean Society of Transportation
• /
• v.26 no.4
• /
• pp.265-273
• /
• 2008

Video Image Processing (VIP) for traffic surveillance has been used not only to gather traffic information, but also to detect traffic conflicts and incident conditions. This paper presents a system development of gathering traffic information and conflict detection based on automatic calculation of pixel length within the detection zone on a Video Detection System (VDS). This algorithm improves the accuracy of traffic information using the automatic detailed line segmentsin the detection zone. This system also can be applied for all types of intersections. The experiments have been conducted with CCTV images, installed at a Bundang intersection, and verified through comparison with a commercial VDS product.

• Journal of the Korean Geotechnical Society
• /
• v.40 no.3
• /
• pp.33-39
• /
• 2024

Inspecting cracks to determine a structure's condition is crucial for accurate safety diagnosis. However, visual crack inspection methods can be subjective and are dependent on field conditions, thereby resulting in low reliability. To address this issue, this study automates the detection of concrete cracks in image data using ResNet, FPN, and the Mask R-CNN components as the backbone, neck, and head of a convolutional neural network. The performance of the proposed model is analyzed using the intersection over the union (IoU). The experimental dataset contained 1,203 images divided into training (70%), validation (20%), and testing (10%) sets. The model achieved an IoU value of 95.83% for testing, and there were no cases where the crack was not detected. These findings demonstrate that the proposed model realized highly accurate detection of concrete cracks in image data.

• Journal of Korea Multimedia Society
• /
• v.18 no.10
• /
• pp.1149-1156
• /
• 2015

This paper proposes the hand contour detector which is robust to noises. Existing methods reduce noises by applying morphology to extracted edges, detect finger tips by using the center of hands, or exploit the intersection of curves from hand area candidates based on J-value segmentation(JSEG). However, these approaches are so vulnerable to noises that are prone to detect non-hand parts. We propose the noise tolerant hand contour detection method in which non-skin area noises are removed by applying skin area detection, contour detection, and a threshold value. By using the implemented system, we observed that the system was successfully able to detect hand contours.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.10
• /
• pp.4176-4197
• /
• 2020

Botnet is a type of dangerous malware. Botnet attack with a collection of bots attacking a similar target and activity pattern is called bot group activities. The detection of bot group activities using intrusion detection models can only detect single bot activities but cannot detect bots' behavioral relation on bot group attack. Detection of bot group activities could help network administrators isolate an activity or access a bot group attacks and determine the relations between bots that can measure the correlation. This paper proposed a new model to measure the similarity between bot activities using the intersections-probability concept to define bot group activities called as B-Corr Model. The B-Corr model consisted of several stages, such as extraction feature from bot activity flows, measurement of intersections between bots, and similarity value production. B-Corr model categorizes similar bots with a similar target to specify bot group activities. To achieve a more comprehensive view, the B-Corr model visualizes the similarity values between bots in the form of a similar bot graph. Furthermore, extensive experiments have been conducted using real botnet datasets with high detection accuracy in various scenarios.

• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.1
• /
• pp.11-18
• /
• 2022

The use of multi-spectral cameras is essential for day and night pedestrian detection. In this paper, a color camera and a thermal imaging infrared camera were used to detect pedestrians near a crosswalk for 24 hours at an intersection with a high risk of traffic accidents. For pedestrian detection, the YOLOv5 object detector was used, and the detection performance was improved by using color images and thermal images at the same time. The proposed system showed a high performance of 0.940 mAP in the day/night multi-spectral (color and thermal image) pedestrian dataset obtained from the actual crosswalk site.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.1
• /
• pp.1-14
• /
• 2014

The expansion of the physical road in response to changes in social conditions and policy of the country has reached the limit. In order to alleviate congestion on the existing road to reconsider the effectiveness of this method should be asking. Currently, how to collect traffic information for management of the intersection is limited to point detection systems. Intelligent Transport Systems (ITS) was the traffic information collection system of point detection method such as through video and loop detector in the past. However, intelligent transportation systems of the next generation(C-ITS) has evolved rapidly in real time interval detection system of collecting various systems between the pedestrian, road, and car. Therefore, this study is designed to evaluate the development of an algorithm for queue length based real-time traffic signal control methodology. Four coordinates estimate on time-space diagram using the travel time each individual vehicle collected via the interval detector. Using the coordinate value estimated during the cycle for estimating the velocity of the shock wave the queue is created. Using the queue length is estimated, and determine the signal timing the total queue length is minimized at intersection. Therefore, in this study, it was confirmed that the calculation of the signal timing of the intersection queue is minimized.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.181-190
• /
• 2022

In this paper, we use a LiDAR sensor and an image camera to detect a left-turning waiting vehicle in two ways, unlike the existing image-type or loop-type left-turn detection system, and a left-turn traffic signal corresponding to the waiting length of the left-turning lane. A system that can efficiently assign a system is introduced. For the LiDAR signal transmitted and received by the LiDAR sensor, the left-turn waiting vehicle is detected in real time, and the image by the video camera is analyzed in real time or at regular intervals, thereby reducing unnecessary computational processing and enabling real-time sensitive processing. As a result of performing a performance test for 5 hours every day for one week with an intersection simulation using an actual signal processor, a detection rate of 99.9%, which was improved by 3% to 5% compared to the existing method, was recorded. The advantage is that 99.9% of vehicles waiting to turn left are detected by the LiDAR sensor, and even if an intentional omission of detection occurs, an immediate response is possible through self-correction using the video, so the excessive waiting time of vehicles waiting to turn left is controlled by all lanes in the intersection. was able to guide the flow of traffic smoothly. In addition, when applied to an intersection in the outskirts of which left-turning vehicles are rare, service reliability and efficiency can be improved by reducing unnecessary signal costs.

• Structural Engineering and Mechanics
• /
• v.39 no.1
• /
• pp.77-91
• /
• 2011

A simply structural damage detection software is developed to identification damage in beams. According to linear fracture mechanics theory, the localized additional flexibility in damage vicinity can be represented by a lumped parameter element. The damaged beam is modeled by wavelet-based elements to gain the first three frequencies precisely. The first three frequencies influencing functions of damage location and depth are approximated by means of surface-fitting techniques to gain damage detection database of forward problem. Then the first three measured natural frequencies are employed as inputs to solve inverse problem and the intersection of the three frequencies contour lines predict the damage location and depth. The DLL (Dynamic Linkable Library) file of damage detection method is coded by C++ and the corresponding interface of software is coded by virtual instrument software LabVIEW. Finally, the software is tested on beams and shafts in engineering. It is shown that the presented software can be used in actual engineering structures.