• 한국재난정보학회 논문집
• /
• 제18권2호
• /
• pp.269-279
• /
• 2022

연구목적: 본 연구의 목적은 건설현장의 재해 예방을 위해 딥러닝기반의 개인보호구 검출 모델을 개발하고, 실제 건설현장에 적용하여 분석하는 것이다. 연구방법: 본 연구의 수행 방법은 실제 환경의 데이터를 구축하고, 개발된 개인보호구 검출 모델을 적용하였다. 개인보호구 검출 모델은 크게 근로자 검출 및 개인보호구 착용 분류 모델로 구성되어 있다. 근로자 검출 모델은 딥러닝 기반의 알고리즘을 실제 현장에서 획득한 데이터셋을 구축하여 학습 및 근로자를 검출하였고, 개인보호구 착용 분류 모델은 앞단에서 추출된 근로자 검출영역에서 학습된 개인보호구 검출 알고리즘을 적용하였다. 구축된 모델의 검증을 위해 건설현장 3곳에서 획득된 데이터를 통해 실험결과를 도출하였다. 연구결과: 데이터베이스 12,000장을 구축하여 정상검출 9,460장(78.8%), 오검출 1,468(12.2%), 미검출 1,072장(8.9%)으로 나타났으며 주요 원인은 영상에서의 객체 크기, 객체간 중첩(Occulusion), 객체 잘림, 그림자에 의한 오검출로 분류되었다. 결론: 개인보호구 검출모델은 현장 상황마다 다른 검출률을 확인할 수 있었고, 본 연구의 결과가 차후 현장적용을 위한 연구에 활용될 수 있을 것으로 여겨진다.

• 산업경영시스템학회지
• /
• 제46권3호
• /
• pp.275-284
• /
• 2023

Land monitoring involves systematically understanding changes in land use, leveraging spatial information such as satellite imagery and aerial photographs. Recently, the integration of deep learning technologies, notably object detection and semantic segmentation, into land monitoring has spurred active research. This study developed a web service to facilitate such integrations, allowing users to analyze aerial and drone images using CNN models. The web service architecture comprises AI, WEB/WAS, and DB servers and employs three primary deep learning models: DeepLab V3, YOLO, and Rotated Mask R-CNN. Specifically, YOLO offers rapid detection capabilities, Rotated Mask R-CNN excels in detecting rotated objects, while DeepLab V3 provides pixel-wise image classification. The performance of these models fluctuates depending on the quantity and quality of the training data. Anticipated to be integrated into the LX Corporation's operational network and the Land-XI system, this service is expected to enhance the accuracy and efficiency of land monitoring.

• 한국지리정보학회지
• /
• 제25권4호
• /
• pp.63-80
• /
• 2022

현실의 공간을 가상의 공간으로 구현하여 문제를 분석하고 예측하는 기술이 개발되면서, 복잡한 도시 내의 정밀한 공간정보를 취득하는 것이 중요해지고 있다. 본 연구는 복잡한 경관을 가진 도시지역을 대상으로 무인항공기를 통해 영상을 취득하고 고해상도 영상에 적합한 영상분류 기법인 객체기반 영상분석 기법과 의미론적 분할 기법을 적용하여 토지피복 분류를 수행하였다. 또한, 동일시기에 수집된 영상을 바탕으로 인공지능이 학습하지 않은 지역에 대해 각 인공지능 모형의 토지피복 분류 재현성을 확인하고자 하였다. 학습 지역을 대상으로 인공지능 모형을 학습하였을 때, 토지피복 분류 정확도가 OBIA-RF는 89.3%, OBIA-DNN은 85.0%, U-Net의 경우 95.3%로 분석되었다. 재현성을 평가하기 위해 검증 지역에 인공지능 모형을 적용하였을 때, OBIA-RF는 7%, OBIA-DNN은 2.1%, U-Net은 2.3%의 정확도가 감소하였다. 형태학적인 특성과 분광학적인 특성을 모두 고려한 U-Net이 토지피복 분류 정확도 및 재현성 평가에서 우수한 성능을 보이는 것으로 나타났다. 본 연구의 결과는 정밀한 공간정보가 중요해짐에 따라 기초자료 생성 방법으로써 도시환경 연구분야에 기여할 수 있을 것으로 판단된다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권8호
• /
• pp.2881-2894
• /
• 2014

In complicated environment, context information plays an important role in image segmentation/labeling. The recently proposed auto-context algorithm is one of the effective context-based methods. However, the standard auto-context approach samples the context locations utilizing a fixed radius sequence, which is sensitive to large scale-change of objects. In this paper, we present a scale invariant auto-context (SIAC) algorithm which is an improved version of the auto-context algorithm. In order to achieve scale-invariance, we try to approximate the optimal scale for the image in an iterative way and adopt the corresponding optimal radius sequence for context location sampling, both in training and testing. In each iteration of the proposed SIAC algorithm, we use the current classification map to estimate the image scale, and the corresponding radius sequence is then used for choosing context locations. The algorithm iteratively updates the classification maps, as well as the image scales, until convergence. We demonstrate the SIAC algorithm on several image segmentation/labeling tasks. The results demonstrate improvement over the standard auto-context algorithm when large scale-change of objects exists.

• 한국인터넷방송통신학회논문지
• /
• 제12권2호
• /
• pp.233-241
• /
• 2012

프로젝트 관리의 목표는 프로젝트를 성공하기 위함이다. 그러나 프로젝트 수행 결과가 성공이라고 명확히 판단할 수 있는가? 또한 판단 결과에 고객과 개발자 모두 동의하는가? 성공과 실패에 대한 정의와 판단 척도들에 대해 다양하게 제시되고 있으나 명확한 분류 기준이 없다. 본 논문은 개발 성공을 결정하기 위한 판단척도들을 살펴보고, 프로젝트의 성공과 실패에 대해 재정의한다. 이러한 정의에 의해 프로젝트 수행 결과를 판단하는 척도와 기준을 제시한다. 제시된 판단 척도와 기준을 적용할 경우, 성공과 실패를 분류하는데 고객과 개발자간의 불협화음을 줄일 수 있을 것이다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권2호
• /
• pp.631-653
• /
• 2020

Sign language is a natural, visually oriented and non-verbal communication channel between people that facilitates communication through facial/bodily expressions, postures and a set of gestures. It is basically used for communication with people who are deaf or hard of hearing. In order to understand such communication quickly and accurately, the design of a successful sign language translation system is considered in this paper. The proposed system includes object detection and classification stages. Firstly, Single Shot Multi Box Detection (SSD) architecture is utilized for hand detection, then a deep learning structure based on the Inception v3 plus Support Vector Machine (SVM) that combines feature extraction and classification stages is proposed to constructively translate the detected hand gestures. A sign language fingerspelling dataset is used for the design of the proposed model. The obtained results and comparative analysis demonstrate the efficiency of using the proposed hybrid structure in sign language translation.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2007년도 추계학술대회논문집
• /
• pp.754-759
• /
• 2007

This study exploits the data discriminating capability of silhouette statistics, which combines wavelet-based vertical energy threshold technique for the purpose of extracting damage-sensitive features and clustering signals of the same class. This threshold technique allows to first obtain a suitable subset of the extracted or modified features of our data, i.e., good predictor sets should contain features that are strongly correlated to the characteristics of the data without considering the classification method used, although each of these features should be as uncorrelated with each other as possible. The silhouette statistics have been used to assess the quality of clustering by measuring how well an object is assigned to its corresponding cluster. We use this concept for the discriminant power function used in this paper. The simulation results of damage detection in a truss structure show that the approach proposed in this study can be successfully applied for locating both open- and breathing-type damage even in the presence of a considerable amount of process and measurement noise.

• 로봇학회논문지
• /
• 제15권4호
• /
• pp.404-412
• /
• 2020

Typically everyday human life tasks involve at least two people moving objects such as tables and beds, and the balancing of such object changes based on one person's action. However, many studies in previous work performed their tasks solely on robots without factoring human cooperation. Therefore, in this paper, we propose cooperative robot for table balancing using Q-learning that enables cooperative work between human and robot. The human's action is recognized in order to balance the table by the proposed robot whose camera takes the image of the table's state, and it performs the table-balancing action according to the recognized human action without high performance equipment. The classification of human action uses a deep learning technology, specifically AlexNet, and has an accuracy of 96.9% over 10-fold cross-validation. The experiment of Q-learning was carried out over 2,000 episodes with 200 trials. The overall results of the proposed Q-learning show that the Q function stably converged at this number of episodes. This stable convergence determined Q-learning policies for the robot actions. Video of the robotic cooperation with human over the table balancing task using the proposed Q-Learning can be found at http://ibot.knu.ac.kr/videocooperation.html.

• 한국농공학회논문집
• /
• 제60권6호
• /
• pp.133-144
• /
• 2018

To suggest the field maintenance plan considering the geographical characteristics of the region, we selected representative regions(plain regione and mountain region) and compared spatial distribution of cultivated land in Muan-gun and Hwasun-gun. Firstly, we examined the distribution characteristics of cultivated land according to the scope of the maintenance object with Fragstats. As a result of that, it was found that the cultivated area except rice paddy had the highest aggregation effect. And then, we developed type classification of maintenance considering geographic characteristics and cultivated crops information. As a result of classification, plain land type Muan region was mostly cultivated land suitable for integrated maintenance. On the other hand, Hwasun, a mountainous terrain, needs small-scale maintenance and road maintenance. Based on these results, it was found that more detailed planning is needed for the upland field infrastructure improvement considering the topographic characteristics.

• Ocean and Polar Research
• /
• 제32권1호
• /
• pp.53-61
• /
• 2010

The distribution of coral reefs can be an indicator of environmental or anthropogenic impacts. Here, we present a habitat map of coral reefs developed using high-spatial satellite images. The study area was located on the north-eastern part of Weno island, in the Chuuk lagoon of Federated States of Micronesia. Two fieldwork expeditions were carried out between 2007 and 2008 to acquire optical and environmental data from 121 stations. We used an IKONOS image obtained in December 2000, and a Kompsat-2 image obtained in September 2008 for the purpose of coral reef mapping. We employed an adapted version of the object-based classification method for efficient classification of the high-spatial satellite images. The habitat map generated using Kompsat-2 was 72.22% accurate in terms of comparative analysis with in-situ measurements. The result of change detection analysis between 2000 and 2008 showed that coral reef distribution had decreased by 6.27% while seagrass meadows had increased by 8.0%.