• Korean Journal of Remote Sensing
• /
• v.39 no.6_1
• /
• pp.1413-1425
• /
• 2023

The increasing frequency of wildfires due to climate change is causing extreme loss of life and property. They cause loss of vegetation and affect ecosystem changes depending on their intensity and occurrence. Ecosystem changes, in turn, affect wildfire occurrence, causing secondary damage. Thus, accurate estimation of the areas affected by wildfires is fundamental. Satellite remote sensing is used for forest fire detection because it can rapidly acquire topographic and meteorological information about the affected area after forest fires. In addition, deep learning algorithms such as convolutional neural networks (CNN) and transformer models show high performance for more accurate monitoring of fire-burnt regions. To date, the application of deep learning models has been limited, and there is a scarcity of reports providing quantitative performance evaluations for practical field utilization. Hence, this study emphasizes a comparative analysis, exploring performance enhancements achieved through both model selection and data design. This study examined deep learning models for detecting wildfire-damaged areas using Landsat 8 satellite images in California. Also, we conducted a comprehensive comparison and analysis of the detection performance of multiple models, such as U-Net and High-Resolution Network-Object Contextual Representation (HRNet-OCR). Wildfire-related spectral indices such as normalized difference vegetation index (NDVI) and normalized burn ratio (NBR) were used as input channels for the deep learning models to reflect the degree of vegetation cover and surface moisture content. As a result, the mean intersection over union (mIoU) was 0.831 for U-Net and 0.848 for HRNet-OCR, showing high segmentation performance. The inclusion of spectral indices alongside the base wavelength bands resulted in increased metric values for all combinations, affirming that the augmentation of input data with spectral indices contributes to the refinement of pixels. This study can be applied to other satellite images to build a recovery strategy for fire-burnt areas.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.89-106
• /
• 2022

Recently, the amount of video data collected from smartphones, CCTVs, black boxes, and high-definition cameras has increased rapidly. According to the increasing video data, the requirements for analysis and utilization are increasing. Due to the lack of skilled manpower to analyze videos in many industries, machine learning and artificial intelligence are actively used to assist manpower. In this situation, the demand for various computer vision technologies such as object detection and tracking, action detection, emotion detection, and Re-ID also increased rapidly. However, the object detection and tracking technology has many difficulties that degrade performance, such as re-appearance after the object's departure from the video recording location, and occlusion. Accordingly, action and emotion detection models based on object detection and tracking models also have difficulties in extracting data for each object. In addition, deep learning architectures consist of various models suffer from performance degradation due to bottlenects and lack of optimization. In this study, we propose an video analysis system consists of YOLOv5 based DeepSORT object tracking model, SlowFast based action recognition model, Torchreid based Re-ID model, and AWS Rekognition which is emotion recognition service. Proposed model uses single-linkage hierarchical clustering based Re-ID and some processing method which maximize hardware throughput. It has higher accuracy than the performance of the re-identification model using simple metrics, near real-time processing performance, and prevents tracking failure due to object departure and re-emergence, occlusion, etc. By continuously linking the action and facial emotion detection results of each object to the same object, it is possible to efficiently analyze videos. The re-identification model extracts a feature vector from the bounding box of object image detected by the object tracking model for each frame, and applies the single-linkage hierarchical clustering from the past frame using the extracted feature vectors to identify the same object that failed to track. Through the above process, it is possible to re-track the same object that has failed to tracking in the case of re-appearance or occlusion after leaving the video location. As a result, action and facial emotion detection results of the newly recognized object due to the tracking fails can be linked to those of the object that appeared in the past. On the other hand, as a way to improve processing performance, we introduce Bounding Box Queue by Object and Feature Queue method that can reduce RAM memory requirements while maximizing GPU memory throughput. Also we introduce the IoF(Intersection over Face) algorithm that allows facial emotion recognized through AWS Rekognition to be linked with object tracking information. The academic significance of this study is that the two-stage re-identification model can have real-time performance even in a high-cost environment that performs action and facial emotion detection according to processing techniques without reducing the accuracy by using simple metrics to achieve real-time performance. The practical implication of this study is that in various industrial fields that require action and facial emotion detection but have many difficulties due to the fails in object tracking can analyze videos effectively through proposed model. Proposed model which has high accuracy of retrace and processing performance can be used in various fields such as intelligent monitoring, observation services and behavioral or psychological analysis services where the integration of tracking information and extracted metadata creates greate industrial and business value. In the future, in order to measure the object tracking performance more precisely, there is a need to conduct an experiment using the MOT Challenge dataset, which is data used by many international conferences. We will investigate the problem that the IoF algorithm cannot solve to develop an additional complementary algorithm. In addition, we plan to conduct additional research to apply this model to various fields' dataset related to intelligent video analysis.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.38 no.3
• /
• pp.249-255
• /
• 2002

This paper describes on the extraction of ship's real-time movement information using a combination full-function ARPA radar and ECS system that displays radar images and an electronic chart together on a single PC screen. The radar target extractor(RTX) board, developed by Marine Electronics Corporation of Korea, receives radar video, trigger, antenna bearing pulse and heading pulse signals from a radar unit and processes these signals to extract target information. The target data extracted from each pulse repetition interval in DSPs of RTX that installed in 16 bit ISA slot of a IBM PC compatible computer is formatted into a series of radar target messages. These messages are then transmitted to the host PC and displayed on a single screen. The position data of target in range and azimuth direction are stored and used for determining the center of the distributed target by arithmetic averaging after the detection of the target end. In this system, the electronic chart or radar screens can be displayed separately or simulaneously and in radar mode all information of radar targets can be recorded and replayed In spite of a PC based radar system, all essential information required for safe and efficient navigation of ship can be provided.

• The Journal of the Acoustical Society of Korea
• /
• v.33 no.4
• /
• pp.262-267
• /
• 2014

Recently, long pulses are transmitted for target detection in active sonar application. Matched filtering implemented by simple convolution algorithm, requires massive computational power for long replica. The computational loads are reduced significantly by implementing the convolution in the frequency domain with overlap add method, but the performance degrades for specified input/output system delay which constrains the size of FFT function. For performance improvement, the replica could be partitioned into uniform blocks (FDL) by re-using IFFT operations, or variable blocks of increasing length (MC) by using the largest possible blocks to calculate the convolution. In this paper, by combining the strong points of the two methods, we propose a new filter partition structure that allows for further optimization of the previous two methods.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.24 no.12
• /
• pp.1190-1197
• /
• 2013

AESA(Active Electronically Scanned Array radar) radar is able to instantaneously and adaptively position and control the beam, and such adaptive beam pointing of AESA radar enables to remarkably improve the multi-mission capability. For this reason, radar resource management(RRM) becomes new challenging issue. RRM is a technique efficiently allocating finite resources, such as energy and time to each task in an optimal and intelligent way. This paper deals with a design of radar resource management algorithms and simulator implemented main algorithms for development of airborne AESA radar. In addition, evaluation results show that developed radar system satisfies a main requirement about simultaneous multiple target tracking and detection by adopting proposed algorithms.

• Ocean and Polar Research
• /
• v.35 no.1
• /
• pp.15-26
• /
• 2013

The toxic, Harmful Algal Blooms (HABs) caused by the Cochlodinium polykrikoides have a serious impact on the coastal waters of Korea. In this study, the acoustic detection system was developed for rapid HABs detection, based on the acoustic backscattering properties of the C. polykrikoides. The developed system was mainly composed of a pulser-receiver board, a signal processor board, a control board, a network board, a power board, ultrasonic sensors (3.5 and 5.0 MHz), an environmental sensor, GPS, and a land-based control unit. To evaluate the performance of the system, a trail was done at a laboratory, and two in situ trials were conducted: (1) when there was no red tide, and (2) when there was red tide. In the laboratory evaluation, the system performed well in accordance with the number of C. polykrikoides in the received level. Second, under the condition when there was no red tide in the field, there was a good correlation between the acoustic data and sampling data. Finally, under the condition when there was red tide in the field, the system successfully worked at various densities in accordance with the number of C. polykrikoides, and the results corresponded with the sampling data and monitoring result of NFRDI (National Fisheries Research & Development Institute). From the laboratory and field evaluations, the developed acoustic detection system for early detecting HABs has demonstrated that it could be a significant system to monitor the occurrence of HABs in coastal regions.

• Journal of Digital Convergence
• /
• v.13 no.4
• /
• pp.235-246
• /
• 2015

Understanding the various unspecified hacking and repeated cyber DDoS attacks, finally was able to find a solution in the methods of attacks. Freely researching approach that combines the attacker and defender, offensive and defensive techniques can be called a challenge to discover the potential in whimsy. In this paper we test and evaluate "KWON-GA", global white hackers team has made by many years of experiences in infiltration and diagnosis under guise of offence is the best defence. And it is knowledge information ICT Convergence security solution which is developed for the purpose of defence, it provide customization policy that can be fit to customer's system environment with needed techniques and it is processed with unique proprietary technology so that it's not possible to scan. And even if it has leaked internally it's impossible to analyze so hackers can't analyze vulnerability, also it can't be abused as hacking tools.

• The Journal of the Korea Contents Association
• /
• v.7 no.3
• /
• pp.93-100
• /
• 2007

Internet had spread in all fields with the fast speed during the last 10 years. Lately, wireless network is also spreading rapidly. Also, number of times that succeed attack attempt and invasion for wireless network is increasing rapidly TMS system was developed to overcome these threat on wireless network. Existing TMS system supplies active confrontation mechanism on these threats. However, existent TMS has limitation that new form of attack do not filtered efficiently. Therefor this paper proposes a new method that it automatically compute the threat from the imput packets with vector space model and detect anomaly detection of wireless network. Proposed mechanism in this research analyzes similarity degree between packets, and detect something wrong symptom of wireless network and then classify these threats automatically.

• Journal of Biomedical Engineering Research
• /
• v.21 no.4
• /
• pp.339-344
• /
• 2000

The mucociliary system is one of the most important airway defense mechanisms in human body and impairment of ciliary movement results in various diseases in respiratory tract. In this study, we have developed a system that can measure ciliary movement in vivo and quantified ciliary beat frequency (CBF) through autoregressive (AR) power spectrum. To measure the frequency in vivo, we applied a photoelectric method that was composed of a laser light and a fiber optic probe. Scattered signals are transferred to a PC in which they are displayed on the monitor and its CBF is determined by the AR method in were acquired. For 8 normal subjects, the analyzed CBFs ranged from 5 to 10Hz and its mean was 7.3${\pm}$1.1Hz. This result showed similar aspects to the reported results of CBFs to data. We expect that this result will be applied in various clinical studies such as analysis of CBF changes by drugs or by diseaes.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.3
• /
• pp.77-85
• /
• 2024

In this dissertation, a we designed and implemented a patrol robot that integrates a thermal imaging camera, speed dome camera, PTZ camera, radar, lidar sensor, and smartphone. This robot has the ability to monitor and respond efficiently even in complex environments, and is especially designed to demonstrate high performance even at night or in low visibility conditions. An orbital movement system was selected for the robot's mobility, and a smartphone-based control system was developed for real-time data processing and decision-making. The combination of various sensors allows the robot to comprehensively perceive the environment and quickly detect hazards. Thermal imaging cameras are used for night surveillance, speed domes and PTZ cameras are used for wide-area monitoring, and radar and LIDAR are used for obstacle detection and avoidance. The smartphone-based control system provides a user-friendly interface. The proposed robot system can be used in various fields such as security, surveillance, and disaster response. Future research should include improving the robot's autonomous patrol algorithm, developing a multi-robot collaboration system, and long-term testing in a real environment. This study is expected to contribute to the development of the field of intelligent surveillance robots.