• 한국멀티미디어학회논문지
• /
• 제25권4호
• /
• pp.559-567
• /
• 2022

The optimal shipping location is determined by measuring the volume and weights of cargo shipped to non-standard cargo carriers. Currently, workers manually measure cargo volume, but automate it to improve work inefficiency. In this paper, we proposed the method of a real-time volume calculation using LiDar sensor for automating cargo measurement of non-standard cargo. For this purpose, we utilized the statistical techniques for data preprocessing and volume calculation, also used Voxel Grid filter to light weighted of data which are appropriate in real-time calculation. We implemented the function of Normal vectors and Triangle Mesh to generate surfaces and Alpha Shapes algorithms to process 3D modeling.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2021년도 봄 학술논문 발표대회
• /
• pp.40-41
• /
• 2021

The construction industry has the highest occupational accidents/injuries among all industries. Korean government installed surveillance camera systems at construction sites to reduce occupational accident rates. Construction safety managers are monitoring potential hazards at the sites through surveillance system; however, the human capability of monitoring surveillance system with their own eyes has critical issues. Therefore, this study proposed to build a deep learning-based safety monitoring system that can obtain information on the recognition, location, identification of workers and heavy equipment in the construction sites by applying multiple-object tracking with instance segmentation. To evaluate the system's performance, we utilized the MS COCO and MOT challenge metrics. These results present that it is optimal for efficiently automating monitoring surveillance system task at construction sites.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2021년도 봄 학술논문 발표대회
• /
• pp.293-294
• /
• 2021

For the successful introduction of the construction automation system, it is necessary to define the consideration factors and target tasks for the necessity of automating the conventional methods. In this study, the factors for automation were derived through brainstorming and questionnaires, and calculated the relative importance through the pairwise comparison using the AHP analysis method. In addition, the detailed type of concrete work was classified into 32 detailed works, and the automation score for each work was calculated by investigating the factors considering the necessity of automation. As a result of the automation score analysis, it was concluded that the need for automation of concrete placing work was the highest. Based on this study, we will consider the strategy of a concrete construction automation system.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2020년도 하계학술대회
• /
• pp.308-310
• /
• 2020

본 논문에서는 포인트 클라우드 정합 시스템 자동화를 위한 재정합 프로세스에서 정합의 실패 유무를 판단하는 기존의 정합 평가 방법을 개선한 방법을 제안한다. 포인트 클라우드 정합 자동화를 위해 정합의 실패를 판단하여 다시 정합하는 재정합 프로세스는 자동화 시스템에서 필수적인 요소이다. 기존의 정합 평가 방법은 정합하고자하는 두 포인트 클라우드의 점의 간격이나 데이터의 양이 다를 경우 계산된 정합 오차가 정성적인 결과와는 다르게 측정되는 문제가 발생하는데, 이는 재정합 프로세스에서 치명적인 오류를 초래한다. 제안하는 방법은 참조 포인트 클라우드에서 가장 인접한 목표 포인트 클라우드의 세 점이 이루는 평면과의 수직 거리를 계산하고, 일정 거리 임계치를 만족하는 점들의 개수를 측정해 계산된 오차를 검증하여 정합 오판단율을 효과적으로 감소시켰다.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.294-298
• /
• 2015

In the United States, the completion of Construction Work Zone (CWZ) impact assessments for all federally-funded highway infrastructure improvement projects is mandated, yet it is regarded as a daunting task for state transportation agencies, due to a lack of standardized analytical methods for developing sounder Transportation Management Plans (TMPs). To circumvent these issues, this study aims to create a spatiotemporal modeling framework, dubbed "SWAT" (Spatiotemporal Work zone Assessment for TMPs). This study drew a total of 43,795 traffic sensor reading data collected from heavily trafficked highways in U.S. metropolitan areas. A multilevel-cluster-driven analysis characterized traffic patterns, while being verified using a measurement system analysis. An artificial neural networks model was created to predict potential 24/7 traffic demand automatically, and its predictive power was statistically validated. It is proposed that the predicted traffic patterns will be then incorporated into a what-if scenario analysis that evaluates the impact of numerous alternative construction plans. This study will yield a breakthrough in automating CWZ impact assessments with the first view of a systematic estimation method.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.10-14
• /
• 2015

Unsafe behavior is one of the major causes of construction accidents. Managing the behavior of workers in real-time is difficult and requires huge manpower. In this paper, a new real-time locating framework is proposed to improve safety management by collecting and analyzing data describing the behavior of workers to identify habits that may lead to accidents. The aim of the study is to identify working habits of workers based on their location history. Location data is used to compare with that of other workers and equipment. The results indicate that the reuse of real-time location data can provide extra safety information for safety management and that the proposed system has the potential to prevent struck-by accidents and caught-in between accidents by predicting unwanted interaction between workers and equipment. This adds to current research aimed at automating construction safety to the point where the continuous monitoring, managing and protection of site workers on site is possible.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.310-312
• /
• 2015

Small Unmanned Aerial Systems (SUAS) have been gaining a special attention in the U.S. recently because it is capable of getting aerial footages conveniently and cost effectively, but also because of its potential threat to the safety of our society. Regarding the benefits, one can easily find successful cases. For example, remote controlled or pre-programmed unmanned aircraft help ranch owners monitor their livestocks or crop harvesting status cost-effectively without having to hire human pilots. The professionals in the construction industry also acknowledge the benefits they could gain from using SUAS. Some firms already use a small unmanned aircraft for monitoring their construction activities, which may help project managers figure out construction progress, resolve disputes in real time, and make proactive decisions for quality control. However, there are many technical challenges that my hinder the use of small unmanned aircraft in the construction industry. This paper explores opportunities and challenges in using unmanned aircraft to monitor concrete cracks on the surface of containment building in the nuclear power plant.

• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 2023년도 제35회 한글 및 한국어 정보처리 학술대회
• /
• pp.190-195
• /
• 2023

의사소통 기술에서 유머는 사람을 웃게 만들며 분위기를 환기시키고, 관계를 돈독하게 만드는 효과를 지닌다. 이를 자연어처리에서 유머 분류, 인식, 탐지로 적용하여 유머를 기계에 학습시키려 하는 다양한 시도가 진행되고 있지만 유머의 주관성과 윤리적 문제로 탁월한 성능을 기록하기 어렵고, 특히 한국어 유머에 대한 자연어처리 분야의 논의는 미비한 상태이다. 이에 본 연구는 유머 평가 체계를 만들어 ChatGPT에 적용하여 유머 인식의 주관성을 극복할 수 있는 자동화 실험을 진행한다. 이때, 유머의 윤리적 문제를 보완하기 위해 한국 법률을 적용한 윤리 기준을 도입하여 유머 데이터셋을 마련하였으며, 데이터셋을 ChatGPT에 fine-tuning 하여 재미있는 생성 모델의 개발 가능성을 실험하였다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2008년도 추계학술발표대회
• /
• pp.766-769
• /
• 2008

디지털 형태의 문서가 널리 퍼지고 끊임없이 증가함에 따라 이를 자동으로 가공하고 처리하는 문서자동분류의 중요성이 널리 인식되고 있다. 본 논문에서는 복합기에서 컬러코드를 인식하는 모듈을 탑재하여 스캔된 문서를 자동으로 분류하는 시스템을 제안하고자 한다. 복합기에서 컬러코드가 부착된 종이문서를 스캔한 다음 그 컬로코드를 추출하여 인식하고 해당 컬러코드와 관련된 문서관리정보에 따라 스캔문서를 복합기 내부의 지정 폴더에 저장하거나 다른 곳으로 전달하는 시스템이다. 이렇게 함으로써 종이문서를 전자화하는 과정에서 수작업으로 분류하는 시간을 줄일 수 있고 또한 사람에 의한 오류를 줄일 수 있다는 장점이 있다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.605-607
• /
• 2023

Recent research has highlighted the effectiveness of Deep Learning (DL) techniques in automating the detection of lung sound anomalies. However, the available lung sound datasets often suffer from limitations in both size and balance, prompting DL methods to employ data preprocessing such as augmentation and transfer learning techniques. These strategies, while valuable, contribute to the increased complexity of DL models and necessitate substantial training memory. In this study, we proposed a streamlined and lightweight DL method but effectively detects lung sound anomalies from small and imbalanced dataset. The utilization of 1D dilated convolutional neural networks enhances sensitivity to lung sound anomalies by efficiently capturing deep temporal features and small variations. We conducted a comprehensive evaluation of the ICBHI dataset and achieved a notable improvement over state-of-the-art results, increasing the average score of sensitivity and specificity metrics by 2.7%.