• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.543-545
• /
• 2019

The AQI index has been developed and used to guide the action of particulate matter. Information on the AQI index can be easily provided to the general public, and various services are provided based on the AQI index. As services are provided, accurate AQI index prediction is needed. In this paper, we design the classification model using the circular neural network to predict the AQI index of particulate matter. For the evaluation of the designed model, compare the AQI index of the actual particulate matter with the predicted value.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.540-542
• /
• 2019

With many announcements on air pollution and human effects from particulate matters, particulate matter forecasts are attracting a lot of public attention. As a result, various efforts have been made to increase the accuracy of particulate matter forecasting by using statistical modeling and machine learning technique. In this paper, the particulate matter AQI index prediction is performed using the multilayer perceptron neural network for particulate matter prediction. For this purpose, a prediction model is designed by using the meteorological factors and particulate matter concentration values commonly used in a number of studies, and the accuracy of the particulate matter AQI prediction is compared.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.1009-1011
• /
• 2022

최근 미세먼지는 중대한 건강위험요소로 고려되고 있고, 미세먼지 취약계층은 이에 대한 적극적 대응이 필요하다. 그러나 현재의 대기환경지수는 세분화 되어있지 않아 본 논문에서는 위해성 평가와 GAM 모형을 기반으로 건강취약계층 대상을 위한 미세먼지 위험지수를 새롭게 개발하였다. 또한, 이에 따라 실내 및 실외활동을 추천하는 시스템을 구현함으로써 생활밀착형 스마트시티로 발돋움하도록 한다.

• Journal of Climate Change Research
• /
• v.9 no.4
• /
• pp.423-433
• /
• 2018

Particulate matter air pollution is a serious problem affecting human health and visibility. The variations in $PM_{10}$ concentrations are influenced by not only local emission sources, but also atmospheric circulation conditions. In this study, we investigate the temporal features of $PM_{10}$ concentrations in South Korea and the atmospheric circulation patterns associated with high concentration episodes of $PM_{10}$ during winter (December-January-February) 2001-2016. Based on those analyses, a Korea Particulate matter Index (KPI) is developed to represent the large-scale atmospheric pattern associated with high concentration episodes of $PM_{10}$. The atmospheric patterns are characterized by persistent high-pressure anomalies, weakened lower-level north-westerly anomalies, and northward shift of the upper-level meridional wind anomalies near the Korean Peninsula. To evaluate the change in occurrence of high concentration episodes of $PM_{10}$ under a possible future warmer climate, we apply KPI analysis to CMIP5 climate simulations. Here, historical and two representative concentration pathway (RCP) scenarios (RCP 4.5 and RCP 8.5) are used. It is found that the occurrence of atmospheric conditions favorable for high $PM_{10}$ concentration episodes tends to increase over South Korea in response to climate change. This suggests that large-scale atmospheric circulation changes under future warmer climate can contribute to increasing high $PM_{10}$ concentration episodes in South Korea.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.1
• /
• pp.55-58
• /
• 2024

In this paper, we propose a design and implementation of big data-based fine dust anomaly detection machine learning system. The proposed is system that classifies the fine dust air quality index through meteorological information composed of fine dust and big data. This system classifies fine dust through the design of an anomaly detection algorithm according to the outliers for each air quality index classification categories based on machine learning. Depth data of the image collected from the camera collects images according to the level of fine dust, and then creates a fine dust visibility mask. And, with a learning-based fingerprinting technique through a mono depth estimation algorithm, the fine dust level is derived by inferring the visibility distance of fine dust collected from the monoscope camera. For experimentation and analysis of this method, after creating learning data by matching the fine dust level data and CCTV image data by region and time, a model is created and tested in a real environment.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.5
• /
• pp.127-133
• /
• 2020

The international cancer research institute under the WHO designated fine dust as a first-class carcinogen. Particular matter refers to dust that is small enough to be invisible and floating in the air. Particular matter is mainly emitted from the combustion process of fossil fuels such as coal and oil, and is a risk factor that can cause lung disease, pneumonia, and heart disease. The Ministry of Environment recently analyzed the output data of 10 fine dust measuring stations and, as a result, announced that about 60% had an error that the existing atmospheric measurement concentration was higher. In order to accurately predict fine dust, the wind direction and measurement position must be corrected. In this paper, in order to solve these problems, fuzzy rules are used to solve these problems. In addition, in order to calculate the fine particulate sensation index actually felt by pedestrians on the street, a computer simulation experiment was conducted to calculate the fine particulate sensation index in consideration of weather conditions, temperature conditions, humidity conditions, and wind conditions.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.409-412
• /
• 2019

국내의 미세먼지 문제가 심각해짐에 따라 대기 오염에 관한 분야의 관심이 높아지고 있다. 현재 정부는 최근 IT 융합 기술의 발전에 따라 빅데이터, 클라우드, 등 사물인터넷 기반 장치의 확산 및 고도화를 위한 기술 접목에 많은 지원과 관심을 보이며 기상청을 통해서는 국내 대기 오염으로 인한 사회적 비용을 낮추기 위해 공공 데이터(Application Program Interface, API)를 활용 다양한 정보 서비스를 지원하고 있다. 하지만 기상청에서 제공하는 정보 서비스에는 한계가 있다. 특히 기상청에서 운영되고 있는 장비들은 고가의 장비로써 비용 및 공간적 설치 제약이 따르며, 약 15km 범위를 한 개소로 담당하여 기상 데이터에 대한 신뢰도에 문제가 발생하고 있다. 본 논문에서는 오픈 하드웨어 기반 소형 기상관측 장비를 활용한 기상지수 및 미세먼지 측정 데이터 제공 시스템을 제안한다. 본 논문에서 제안한 시스템은 기상 계측이 필요한 지역의 작은 공간을 활용, 기상관측 장비를 통해 관측된 데이터와 기상청에서 제공하는 생활 기상지수 알고리즘을 토대로 해당 지역에 맞는 맞춤형 정보를 제공하여 사회적 비용을 낮출 수 있을 것으로 기대한다.

• Korean Journal of Optics and Photonics
• /
• v.31 no.2
• /
• pp.71-80
• /
• 2020

In this article, we suggest a new method for measuring an aerosol's extinction coefficient using a commercial camera. For a given image, we choose three pixel-points that are imaged for the same kinds of objects located in similar directions. We suggest and calculate aerosol extinction coefficients from these RGB gray levels and the different distances of the three objects. To compare our measurement results, we also measure extinction coefficients using lidar. Finally, we find that there are meaningful and sensible correlations between these two measurements, with a correlation coefficient of 0.86. We measure the aerosol extinction coefficient at three different RGB wavelengths using the same method. From these aerosol extinction coefficients at three different wavelengths, we find that the Angstrom exponent ranges from 0.7 to 1.6 over a full daytime period. We believe that these Angstrom exponents can give important information about the size of the fine particles.

• Air Cleaning Technology
• /
• v.33 no.3
• /
• pp.37-43
• /
• 2020

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.6
• /
• pp.573-581
• /
• 2020

Using CCTV images and weather parameters, a method for estimating PM (Particulate Matter) index was proposed, and an experiment was conducted. For CCTV images, we proposed a method of estimating the PM index by applying a deep learning technique based on a CNN (Convolutional Neural Network) with ROI(Region Of Interest) image including a specific spot and an full area image. In addition, after combining the predicted result values by deep learning with the two weather parameters of humidity and wind speed, a post-processing experiment was also conducted to calculate the modified PM index using the learned regression model. As a result of the experiment, the estimated value of the PM index from the CCTV image was R2(R-Squared) 0.58~0.89, and the result of learning the ROI image and the full area image with the measuring device was the best. The result of post-processing using weather parameters did not always show improvement in accuracy in all cases in the experimental area.