• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권2호
• /
• pp.381-404
• /
• 2022

Forest fires inflict great losses of human lives and serious damages to ecological systems. Hence, numerous fire detection methods have been proposed, one of which is fire detection based on sensors. However, these methods reveal several limitations when applied in large spaces like forests such as high cost, high level of false alarm, limited battery capacity, and other problems. In this research, we propose a novel forest fire detection method based on image processing and correlation coefficient. Firstly, two fire detection conditions are applied in RGB color space to distinguish between fire pixels and the background. Secondly, the image is converted from RGB to YCbCr color space with two fire detection conditions being applied in this color space. Finally, the correlation coefficient is used to distinguish between fires and objects with fire-like colors. Our proposed algorithm is tested and evaluated on eleven fire and non-fire videos collected from the internet and achieves up to 95.87% and 97.89% of F-score and accuracy respectively in performance evaluation.

• 한국환경과학회지
• /
• 제31권9호
• /
• pp.735-741
• /
• 2022

Global warming affects forests and their ecology. Diversity in the forest is a buffer that reduces the damage due to global warming. Mixed forests are ecologically more valuable as versatile habitats and are effective in preventing landslides. In Korea, most forests were created by simple afforestation with trees of evergreen species. Typically, evergreen trees are shallow-rooted, and deciduous trees are deep-rooted. Mixed forest tree roots grip the soil effectively, which reduces the occurrence of landslides. Therefore, improving the distribution of tree types is essential to reduce damage due to global warming. For this improvement, the investigation of tree types of the forest is needed. However, determining the tree type distribution of forests that are spread over wide areas is labor-intensive and time-consuming. This study suggests effective methods for determining the distribution of tree types in a forest that is spread across a relatively wide area. Using normalized difference vegetation index and RGB images from unmanned aerial vehicles, each evergreen and deciduous tree, and grassland area can be distinguished. The distinguished image determines the distribution of tree type. This method is effective compared to directly determining the tree type distribution in the forest by the use of manpower. The data from these methods could be applied to plan a mixed forest or to prepare for future damage due to global warming.

• Smart Structures and Systems
• /
• 제31권4호
• /
• pp.351-363
• /
• 2023

Visual structural inspections are an inseparable part of post-earthquake damage assessments. With unmanned aerial vehicles (UAVs) establishing a new frontier in visual inspections, there are major computational challenges in processing the collected massive amounts of high-resolution visual data. We propose twin deep learning models that can provide accurate high-resolution structural components and damage segmentation masks efficiently. The traditional approach to cope with high memory computational demands is to either uniformly downsample the raw images at the price of losing fine local details or cropping smaller parts of the images leading to a loss of global contextual information. Therefore, our twin models comprising Trainable Resizing for high-resolution Segmentation Network (TRS-Net) and DmgFormer approaches the global and local semantics from different perspectives. TRS-Net is a compound, high-resolution segmentation architecture equipped with learnable downsampler and upsampler modules to minimize information loss for optimal performance and efficiency. DmgFormer utilizes a transformer backbone and a convolutional decoder head with skip connections on a grid of crops aiming for high precision learning without downsizing. An augmented inference technique is used to boost performance further and reduce the possible loss of context due to grid cropping. Comprehensive experiments have been performed on the 3D physics-based graphics models (PBGMs) synthetic environments in the QuakeCity dataset. The proposed framework is evaluated using several metrics on three segmentation tasks: component type, component damage state, and global damage (crack, rebar, spalling). The models were developed as part of the 2^nd International Competition for Structural Health Monitoring.

• 한국항행학회논문지
• /
• 제27권4호
• /
• pp.473-480
• /
• 2023

본 연구에서는 손 제스처를 이용하여 탐색 구조용 무인항공기의 카메라 스캐닝 경로를 입력하는 방법을 개발하여 가상현실 환경에서 시험하였다. 두 개의 임무영역 - 화재 발생 빌딩과 산악지역-에서 인명구조 탐색을 위해 관심 지역 빌딩과 산악 영역의 축소형 디지털트윈을 각각의 관심 지역 앞에 생성하고 디지털트윈 표면 상에 손 제스처로 비행경로를 정의한 후 무인항공기의 자동비행이 이루어지도록 하였다. 19명의 실험참가자로 이루어진 Human-in-the-loop 시뮬레이션 테스트에서 기존의 매뉴얼 비행 방식과 이 연구에서 설계한 경로 정의 후 자동비행 방식을 비교하였다. 테스트 결과, 객관적 결과는 새로운 방식이 임무의 정확성을 전반적으로 더 높였다고 보기는 어려웠으나, 주관적 답변은 산악지역 임무시 더 낮은 작업부하를 도출한다는 것을 보여주었다. 이러한 전문가의 특수한 업무능력에 의존하는 탐색구조 분야에서는 업무부하 감소가 필요한 부분에 선택적으로 적용하여 자동화로 보완해주는 것이 필요하다.

• Journal of Animal Science and Technology
• /
• 제65권3호
• /
• pp.638-651
• /
• 2023

The objective of this study was to quantitatively estimate the level of grazing area damage in outdoor free-range pig production using a Unmanned Aerial Vehicles (UAV) with an RGB image sensor. Ten corn field images were captured by a UAV over approximately two weeks, during which gestating sows were allowed to graze freely on the corn field measuring 100 × 50 m². The images were corrected to a bird's-eye view, and then divided into 32 segments and sequentially inputted into the YOLOv4 detector to detect the corn images according to their condition. The 43 raw training images selected randomly out of 320 segmented images were flipped to create 86 images, and then these images were further augmented by rotating them in 5-degree increments to create a total of 6,192 images. The increased 6,192 images are further augmented by applying three random color transformations to each image, resulting in 24,768 datasets. The occupancy rate of corn in the field was estimated efficiently using You Only Look Once (YOLO). As of the first day of observation (day 2), it was evident that almost all the corn had disappeared by the ninth day. When grazing 20 sows in a 50 × 100 m² cornfield (250 m²/sow), it appears that the animals should be rotated to other grazing areas to protect the cover crop after at least five days. In agricultural technology, most of the research using machine and deep learning is related to the detection of fruits and pests, and research on other application fields is needed. In addition, large-scale image data collected by experts in the field are required as training data to apply deep learning. If the data required for deep learning is insufficient, a large number of data augmentation is required.

• 한국작물학회:학술대회논문집
• /
• 한국작물학회 2022년도 추계학술대회
• /
• pp.75-75
• /
• 2022

Traditional agriculture mostly focused on activity in the field, but current agriculture faces problems such as reduction of agricultural inputs, labor shortage and so on. Accordingly, traditional agricultural experiments generally considered the simple treatment effects, but current agricultural experiments need to consider the several and complicate treatment effects. To analyze such several and complicate treatment effects, data collection has the first priority. Remote sensing is a quite effective tool to collect information in agriculture, and recent easier availability of UAVs (Unmanned Aerial Vehicles) enhances the effectiveness. LAI (Leaf Area Index) is one of the most important information for evaluating the condition of crop growth. In this study, we utilized UAV with multispectral camera to evaluate plant-based LAI of sweetcorn in a small-scale field experiment and discussed the feasibility of a new experimental design to analyze the several and complicate treatment effects. The plant-based SR measured by UAV showed the highest correlation coefficient with LAI measured by a canopy analyzer in 2018 and 2019. Application of linear mix model showed that plant-based SR data had higher detection power due to its huge number of data although SR was inferior to evaluate LAI than the canopy analyzer. The distribution of plant-based data also statistically revealed the border effect in treatment plots in the traditional experimental design. These results suggest that remote sensing with UAVs has the advantage even in a small-scale experimental plot and has a possibility to provide a new experimental design if combined with various analytical applications such as plant size, shape, and color.

• 대한원격탐사학회지
• /
• 제40권1호
• /
• pp.9-18
• /
• 2024

Small streams, despite their rich ecosystems, face challenges in vegetation assessment due to the limitations of traditional, time-consuming methods. This study presents a groundbreaking approach, combining unmanned aerial vehicles(UAVs), convolutional neural networks(CNNs), and the vegetation differential vegetation index (VDVI), to revolutionize both assessment and management of stream vegetation. Focusing on Idong Stream in South Korea (2.7 km long, 2.34 km² basin area)with eight diverse revetment methods, we leveraged high-resolution RGB images captured by UAVs across five dates (July-December). These images trained a ResNeXt101 CNN model, achieving an impressive 89% accuracy in classifying vegetation cover(soil,water, and vegetation). This enabled detailed spatial and temporal analysis of vegetation distribution. Further, VDVI calculations on classified vegetation areas allowed assessment of vegetation vitality. Our key findings showcase the power of this approach:(a) TheCNN model generated highly accurate cover maps, facilitating precise monitoring of vegetation changes overtime and space. (b) August displayed the highest average VDVI(0.24), indicating peak vegetation growth crucial for stabilizing streambanks and resisting flow. (c) Different revetment methods impacted vegetation vitality. Fieldstone sections exhibited initial high vitality followed by decline due to leaf browning. Block-type sections and the control group showed a gradual decline after peak growth. Interestingly, the "H environment block" exhibited minimal change, suggesting potential benefits for specific ecological functions.(d) Despite initial differences, all sections converged in vegetation distribution trends after 15 years due to the influence of surrounding vegetation. This study demonstrates the immense potential of UAV-based remote sensing and CNNs for revolutionizing small-stream vegetation assessment and management. By providing high-resolution, temporally detailed data, this approach offers distinct advantages over traditional methods, ultimately benefiting both the environment and surrounding communities through informed decision-making for improved stream health and ecological conservation.

• 항공우주시스템공학회지
• /
• 제17권6호
• /
• pp.16-26
• /
• 2023

UAV의 수요가 증가함에 따라 많은 충돌 회피 방법들이 제안됐다. 이러한 방법들은 LiDAR 및 스테레오 카메라를 주축으로 연구되었으나 무겁거나 공간이 부족하여 소형 UAV에 접목이 어려웠기에, 최근에는 객체 인지 모델 및 거리 측정 센서를 복합적으로 사용한 방법들이 제안되고 있다. 하지만 이러한 객체 인지 복합 방법들은 인지한 장애물의 크기 정보를 도출하지 않아 인지 초기에 적정 회피 거리 도출 및 장애물의 좌표화가 어렵다는 단점이 존재한다. 본 논문에서는 단안 카메라-YOLO와 적외선 센서 기반의 장애물 크기 예측 방법을 제안하고, 실험을 통해 40cm의 거리 내에서 86.39%의 정확도를 보임을 확인했다. 또한, 제안한 방법을 적용하여 소형 UAV에 적용하여 장애물 충돌 회피가 가능한지를 확인하였다.

• 항공우주시스템공학회지
• /
• 제18권3호
• /
• pp.48-55
• /
• 2024

최근에 무인항공기의 사업화가 활발하게 추진됨에 따라 무인항공기의 안전성 확보를 위한 기술 개발에 많은 관심이 집중되고 있다. 일반적으로 무인항공기는 운용 중 급기동, 외란, 조종사 실수 등으로 인하여 조종 불능의 상태로 진입할 가능성을 지닌다. 조종 불능 상태로의 진입을 예방하기 위해서는 무인항공기의 비행 상태를 예측하는 것이 필수적으로 요구된다. 본 논문에서는 무인항공기의 비행 상태 예측 성능의 향상을 위하여 개선된 CNN-LSTM 혼합모델을 제안한다. 모의실험은 제안하는 모델을 이용한 예측 기법이 기존 예측 기법에 비하여 비행 상태 예측 성능이 우수하며 온보드 환경에서 실시간으로 운용됨을 보인다.

• 항공우주시스템공학회지
• /
• 제18권3호
• /
• pp.41-47
• /
• 2024

최근에 무인항공기의 실용화 및 사업화가 추진됨에 따라 무인항공기의 안전성 확보에 관한 관심이 증가하고 있다. 무인항공기의 사고는 재산 및 인명 피해를 발생시키기 때문에 사고를 예방할 수 있는 기술의 개발은 중요하다. 이러한 이유로 AutoEncoder 모델을 이용한 비정상 비행 상태 탐지 기법이 개발되었다. 그러나 기존 탐지 기법은 성능과 실시간 처리 측면에서 한계를 지닌다. 본 논문에서는 U-Net 기반 비정상 비행 탐지 기법을 제안한다. 제안하는 기법에서는 U-Net 모델에서 얻어지는 재구성 오차에 대한 마할라노비스 거리 증가량에 기반하여 비정상 비행이 탐지된다. 모의실험을 통해 제안 탐지 기법이 기존 탐지 기법에 비해 탐지 성능이 우수하며 온보드 환경에서 실시간으로 구동될 수 있음을 알 수 있다.