바로가기메뉴

본문 바로가기 주메뉴 바로가기

logo

Search Word: Unmanned aerial vehicle, Search Result: 2
1
Ho-Gyeong Moon(Team of Cooperation on Wetlands, National Institute of Ecology) ; Han Kim(Team of Cooperation on Wetlands, National Institute of Ecology) ; Nak-Hyun Choi(Team of Cooperation on Wetlands, National Institute of Ecology) ; Dong-Pil Kim(Department of Landscape Architecture, Pusan National University) 2020, Vol.1, No.1, pp.31-40 https://doi.org/10.22920/PNIE.2020.1.1.31
초록보기
Abstract

The rapid development of technologies in unmanned aerial vehicles (UAVs) has led to their use in various areas. UAVs are mainly used for commercial purposes, but their utilization is increasingly important in other areas because their operation cost is less than satellites and aerial imaging. The utilization of UAVs in the environment/ecology area is relatively new. Therefore, identifying the trends of UAV-related spatial information is significant in basic research for UAV utilization. This study quantitatively identified domestic and international research trends related to UAV utilization and analyzed research areas. An attempt was also made to identify upcoming UAV-related topics in the environment/ecology research field using text mining to analyze the bibliographic information of global research literature. Domestic UAV-related studies were classified into seven clusters where basic research on “UAV technology/industry trends” was abundant, and studies on data collection and analysis through UAV remote sensing technology have increased since 2015. Eight clusters were identified for international studies where the most active research area international was “remote sensing technology/data analysis”. In addition, Canopy, Classification, Forest, Leaf Area Index, Normalized Difference Vegetation Index, Temperature, Tree, and Atmosphere appeared as the main keywords related to environment and ecology. The appearance frequencies and association strengths were high because the advancement in UAV optical sensor technology and the rapid development of image processing technology enabled the acquisition of data that could not be obtained from existing spatial information. They are recognized as future research topics as related domestic studies have begun corresponding to international research.


2
Deokjin Joo(Hashed) ; Jungmin You(Research Institute of Ecoscience, Ewha Womans University) ; Yong-Jin Won(Division of EcoScience, Ewha Womans University) 2022, Vol.3, No.2, pp.67-72 https://doi.org/10.22920/PNIE.2022.3.2.67
초록보기
Abstract

Ecological research relies on the interpretation of large amounts of visual data obtained from extensive wildlife surveys, but such large-scale image interpretation is costly and time-consuming. Using an artificial intelligence (AI) machine learning model, especially convolution neural networks (CNN), it is possible to streamline these manual tasks on image information and to protect wildlife and record and predict behavior. Ecological research using deep- learning-based object recognition technology includes various research purposes such as identifying, detecting, and identifying species of wild animals, and identification of the location of poachers in real-time. These advances in the application of AI technology can enable efficient management of endangered wildlife, animal detection in various environments, and real-time analysis of image information collected by unmanned aerial vehicles. Furthermore, the need for school education and social use on biodiversity and environmental issues using AI is raised. School education and citizen science related to ecological activities using AI technology can enhance environmental awareness, and strengthen more knowledge and problem-solving skills in science and research processes. Under these prospects, in this paper, we compare the results of our early 2013 study, which automatically identified African cichlid fish species using photographic data of them, with the results of reanalysis by CNN deep learning method. By using PyTorch and PyTorch Lightning frameworks, we achieve an accuracy of 82.54% and an F1-score of 0.77 with minimal programming and data preprocessing effort. This is a significant improvement over the previous our machine learning methods, which required heavy feature engineering costs and had 78% accuracy.

Proceedings of the National Institute of Ecology of the Republic of Korea