• Korean Journal of Artificial Intelligence
• /
• v.11 no.3
• /
• pp.17-22
• /
• 2023

In this paper, we propose a black ice detection system using Convolutional Neural Networks (CNNs). Black ice poses a serious threat to road safety, particularly during winter conditions. To overcome this problem, we introduce a CNN-based architecture for real-time black ice detection with an encoder-decoder network, specifically designed for real-time black ice detection using thermal images. To train the network, we establish a specialized experimental platform to capture thermal images of various black ice formations on diverse road surfaces, including cement and asphalt. This enables us to curate a comprehensive dataset of thermal road black ice images for a training and evaluation purpose. Additionally, in order to enhance the accuracy of black ice detection, we propose a multi-scale dilation convolution feature fusion (MsDC-FF) technique. This proposed technique dynamically adjusts the dilation ratios based on the input image's resolution, improving the network's ability to capture fine-grained details. Experimental results demonstrate the superior performance of our proposed network model compared to conventional image segmentation models. Our model achieved an mIoU of 95.93%, while LinkNet achieved an mIoU of 95.39%. Therefore, it is concluded that the proposed model in this paper could offer a promising solution for real-time black ice detection, thereby enhancing road safety during winter conditions.

• Korean Journal of Veterinary Service
• /
• v.16 no.2
• /
• pp.111-119
• /
• 1993

In this study, we carried out the rapid diagnostic system based on indirect fluorescent anti-body technique (IFAT) for detection of bacterial diseases in cultured freshwater fishes. 1. When the fishes were tested with graded dilution of Edwardsiella tarda FPC 470 bacteria detection from ten fishes Injected with $4.1{\times}10^3$colony forming unit(CFU) /ml, all of them were detected by IFAT but only two fishes were recognizable by the culture method in the tested fishes injected with $4.1{\times}10^3$CFU /ml. 2. The bacteria E. tarda could be detected by IFAT method from 1 to 48hrs after Injection in the tissues tested such as kidney, liver and spleen of the fishes, whereas detection by culture method could be recognized from 1 to 48hrs after injection In the kidney and spleen but it was not possible from preinjection to 1 hr in the liver. 3. Thus, IFAT proved to be more useful technique than plate culture method in the diagnosis of Edwardsiellosis in the freshwater fishes.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.2
• /
• pp.86-92
• /
• 2001

Various array processing techniques to identify the noise source position or bearing have been developed. Typical array processing techniques which are based on time delay between received signals at two sensors, are classified as conventional beamforming, correlation function and NAH (Near-Field Acoustic Holography) techniques which have their own characteristics with respect to application field and signal processing method. In this study, correlation function technique which could be applied for broadband noise source detection, is adopted and the effective detection technique is proposed considering the effects of source bandwidth and measurement noise correlation of noise sources. The validity of the Proposed technique is evaluated using the 3-dimensional nonlinear any which does not give 3-dimensional Position or bearing ambiguity

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.6
• /
• pp.995-1000
• /
• 2017

In this paper, we present an electrical arc detection algorithm using the mash-up analysis technique which is the core technology for the autonomous electrical safety management system(AESMS) of the multi-unit dwellings. The mash-up analysis technique analyzes the voltage, load current, zero phase current data simultaneously to judge arc faults. In order to develop the arc fault detection algorithm, the characteristics of series arc and parallel arc were analyzed. Also, we propose the mash-up analysis technique that analyzes waveforms of voltage, load current, and zero phase current at the same time. The arc fault detection algorithm was developed using the mash-up analysis technique. The developed algorithm can prevent electrical disasters in an effective way through accident prediction, and it will be used as a basic technology to introduce an autonomous electrical safety management system.

• Wind and Structures
• /
• v.21 no.6
• /
• pp.709-726
• /
• 2015

In this study, a novel vision-based bolt-loosening monitoring technique is proposed for bolted joints connecting tubular steel segments of the wind turbine tower (WTT) structure. Firstly, a bolt-loosening detection algorithm based on image processing techniques is developed. The algorithm consists of five steps: image acquisition, segmentation of each nut, line detection of each nut, nut angle estimation, and bolt-loosening detection. Secondly, experimental tests are conducted on a lab-scale bolted joint model under various bolt-loosening scenarios. The bolted joint model, which is consisted of a ring flange and 32 sets of bolt and nut, is used for simulating the real bolted joint connecting steel tower segments in the WTT. Finally, the feasibility of the proposed vision-based technique is evaluated by bolt-loosening monitoring in the lab-scale bolted joint model.

• International Journal of Aeronautical and Space Sciences
• /
• v.7 no.1
• /
• pp.73-83
• /
• 2006

In this paper, a redundancy management system for aircraft is studied, and fault detection and isolation algorithms of inertial sensor system are proposed. Contrary to the conventional aircraft systems, UAV system cannot allow triple or quadruple hardware redundancy due to the limitations on space and weight. In the UAV system with dual sensors, it is very difficult to identify the faulty sensor. Also, conventional fault detection and isolation (FDI) method cannot isolate multiple faults in a triple redundancy system. In this paper, two FDI techniques are proposed. First, hardware based FDI technique is proposed, which combines a parity equation approach with a wavelet based technique. Second, analytic FDI technique based on the Kalman filter is proposed, which is a model-based FDI method utilizing the threshold value and the confirmation time. To provide the reference value for detecting the fault, residuals are calculated using the extended Kalman filter. To verify the effectiveness of the proposed FDI methods, numerical simulations are performed.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.3
• /
• pp.187-192
• /
• 2006

We reviewed the collision event with pc-based shooting games and existing collision detection algorithms. Then we proposed a new collision check technique using a small quadrilateral unit. by which existing quadrilateral collision detection techniques can be made up for. For demonstration we implemented a simple shooting game having its screen output. We proved that the proposed technique can be applied to real computer games by means of showing the experimental results and screen outputs from implemented real games.

• Smart Media Journal
• /
• v.7 no.4
• /
• pp.79-89
• /
• 2018

In recent years, falls among elderly people have gained serious attention as a major cause of injuries. Falls often lead to fatal consequences due to lack of prompt response and rescue. Therefore, a more accurate fall detection system and an effective feature extraction technique are required to prevent and reduce the risk of such incidents. In this paper, we proposed an efficient feature extraction technique based on multiple sliding windows and validated it through a series of experiments using supervised learning algorithms. The experiments were conducted using the public datasets obtained from tri-axial accelerometers. The results depicted that extraction of the feature from adjacent sliding windows led to high accuracy in supervised machine learning-based fall detection. Also, the experiments conducted in this study suggested that the best accuracy can be achieved by keeping the window size as small as 2 seconds. With the kNN classifier and dataset from wearable sensors, the experiments achieved accuracy rates of 94%.

• Journal of Navigation and Port Research
• /
• v.37 no.5
• /
• pp.527-533
• /
• 2013

This paper presents damage detection and estimation of stiffness parameter on a concrete scale model and a real structure of concrete pontoon using dynamic properties such as mode shapes and natural frequencies. In case of damage detection, dynamic impact test on a concrete scale model is accomplished to extract mode shapes and the practicality is verified by utilizing a damage detection technique. And the stiffness parameter of a real structure of concrete pontoon was estimated via system identification technique using the natural frequencies of the structure. The results indicate that the damaged elements of the scale model are found exactly using damage detection technique and the effective stiffness property of the real structure of concrete pontoon can be estimated by system identification technique.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.05a
• /
• pp.411-414
• /
• 2009

Many victims and property damages are caused in fires every year. In this paper, intelligent fire-detection systems with embedded fire-detection algorithms for early fire detection and alarm is proposed to reduce fire damages by using image processing technique, high speed digital signal processor(DSP) technique, and information technique. The fire detection algorithms used for the proposed systems consist of flame and smoke detection algorithms. If flame or smoke is detected respectively, the corresponding alarm signal can be transferred to management computer. And if flame and smoke is detected simultaneously, the fire alarm signal shall be generated. Through several experiments in the physical environment, it is shown that the proposed system works well without malfunction.