http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Deep learning-based oil spill detection with LWIR camera
박호민,박경수,김정근,김재훈,이성대 한국마린엔지니어링학회 2021 한국마린엔지니어링학회지 Vol.45 No.6
The ocean covers approximately 71% of the total surface area of Earth and plays a significant role in maintaining the envi-ronment and ecosystems. Oil spills are the largest source of pollution in the ocean, mainly Bunker C oil and diesel oil used as vessel fuels. Therefore, oil spill detection is essential for marine protection, which motivated this study. Detection with radar, which is based on electromagnetic waves, is achieved using satellite synthetic aperture radar (SAR); thus, real-time detection over a small range is difficult. Hence, in this study, an oil spill detection system based on thermal imaging using a long-wave infrared (LWIR) camera is proposed. The oil spill detection algorithm utilizes the You Only Look Once (YOLO) model, which is widely used for object detection. In addition, 1,644 thermal images were labeled to evaluate the proposed system. The training and test results showed an accuracy of 96.91% and false alarm rate of 8.33%. An improved detection performance can be expected from subsequent experiments using larger image datasets.
Deep learning–based drone detection with SWIR cameras
박호민,박경수,김유린,김정근,김재훈,이성대 한국마린엔지니어링학회 2020 한국마린엔지니어링학회지 Vol.44 No.6
Small unmanned aerial vehicles, commonly known as drones, and their related industries are improving in leaps and bounds. The global drone industry began with a military focus and subsequently progressed into commercial applications. Consequently, abuse cases linked to drone technology are gradually increasing. Following the technical advancement in drone technology, studies on drone detection and prevention are actively ongoing. This is one such study. Radar-based drone detection that combines various existing sensors or equipment has shortcomings, including high costs and specialist operations. Thus, this paper proposes a drone-detection system that uses only thermal images from short-wavelength infrared (SWIR) cameras. The YOLO model, which is widely used for object recognition, was used for the drone-detection algorithm. Labels were attached to 22,921 thermal images to test the constructed system; 16,121 images were used for training and the remainder for testing. The test results showed 98.17% precision and 98.65% recall. Learning through drone-image shooting in various environments, after removing static from clouds and other noise, is expected to improve detection performance in the future.
박호민,김창현,김재훈 한국정보처리학회 2020 정보처리학회논문지. 소프트웨어 및 데이터 공학 Vol.9 No.2
Sentiment analysis is the automated process of understanding attitudes and opinions about a given topic from written or spoken text. One of the sentiment analysis approaches is a dictionary-based approach, in which a sentiment dictionary plays an much important role. In this paper, we propose a method to automatically generate Korean sentiment lexicon from the well-known English sentiment lexicon called VADER (Valence Aware Dictionary and sEntiment Reasoner). The proposed method consists of three steps. The first step is to build a Korean-English bilingual lexicon using a Korean-English parallel corpus. The bilingual lexicon is a set of pairs between VADER sentiment words and Korean morphemes as candidates of Korean sentiment words. The second step is to construct a bilingual words graph using the bilingual lexicon. The third step is to run the label propagation algorithm throughout the bilingual graph. Finally a new Korean sentiment lexicon is generated by repeatedly applying the propagation algorithm until the values of all vertices converge. Empirically, the dictionary-based sentiment classifier using the Korean sentiment lexicon outperforms machine learning-based approaches on the KMU sentiment corpus and the Naver sentiment corpus. In the future, we will apply the proposed approach to generate multilingual sentiment lexica. 감정분석은 문서 또는 대화상에서 주어진 주제에 대한 태도와 의견을 이해하는 과정이다. 감정분석에는 다양한 접근법이 있다. 그 중 하나는 감정사전을 이용하는 사전 기반 접근법이다. 본 논문에서는 널리 알려진 영어 감정사전인 VADER를 활용하여 한국어 감정사전을 자동으로 생성하는 방법을 제안한다. 제안된 방법은 세 단계로 구성된다. 첫 번째 단계는 한영 병렬 말뭉치를 사용하여 한영 이중언어 사전을 제작한다. 제작된 이중언어 사전은 VADER 감정어와 한국어 형태소 쌍들의 집합이다. 두 번째 단계는 그 이중언어 사전을 사용하여 한영 단어 그래프를 생성한다. 세 번째 단계는 생성된 단어 그래프 상에서 레이블 전파 알고리즘을 실행하여 새로운 감정사전을 구축한다. 이와 같은 과정으로 생성된 한국어 감정사전을 유용성을 보이려고 몇 가지 실험을 수행하였다. 본 논문에서 생성된 감정사전을 이용한 감정 분류기가 기존의 기계학습 기반 감정분류기보다 좋은 성능을 보였다. 앞으로 본 논문에서 제안된 방법을 적용하여 여러 언어의 감정사전을 생성하려고 한다.
항공기 형상에 대한 근전계 RCS 측정에서 내삽 알고리즘을 이용한 측정시간 단축에 대한 분석
박호민 한국군사과학기술학회 2022 한국군사과학기술학회지 Vol.25 No.4
The importance of stealth technology is increasing in modern warfare, and Radar Cross Section(RCS) is widely used as an indicator of stealth technology. It is useful to measure RCS using an image-based near-field to far-field transformation algorithm in short-range monostatic conditions. However, the near-field measurement system requires a longer measurement time compared to other methods. In this work, it is proposed to reduce the measured data using an interpolation method in azimuth angular domain. The calculated far-field RCS values according to the sampling rate is shown, and the performance of the algorithm applied with interpolation in the angular domain is presented. It is shown that measurement samples can be reduced several times by using the redundancy in the angular domain while producing results similar to the conventional method.