• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제27권4호
• /
• pp.311-323
• /
• 2023

This paper introduces the industrial problem, the solution, and the results of the research conducted with Define Inc. The client company wanted to improve the performance of an object detection model on the floor plan dataset. To solve the problem, we analyzed the operational principles, advantages, and disadvantages of the existing object detection model, identified the characteristics of the floor plan dataset, and proposed to use of YOLO v5 as an appropriate object detection model for training the dataset. We compared the performance of the existing model and the proposed model using mAP@60, and verified the object detection results with real test data, and found that the performance increase of mAP@60 was 0.08 higher with a 25% shorter inference time. We also found that the training time of the proposed YOLO v5 was 71% shorter than the existing model because it has a simpler structure. In this paper, we have shown that the object detection model for the floor plan dataset can achieve better performance while reducing the training time. We expect that it will be useful for solving other industrial problems related to object detection in the future. We also believe that this result can be extended to study object recognition in 3D floor plan dataset.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1991년도 한국자동제어학술회의논문집(국내학술편); KOEX, Seoul; 22-24 Oct. 1991
• /
• pp.663-668
• /
• 1991

In order to navigate, mobile robot needs to avoid obstacles on his way. We describe a stereo vision method for detecting obstacles on the floor ground. With the knowledge of floor geometry, stereo images are transformed so that the relative views of obstacle to the floor are seen. After comparing the transformed images, obstacles information such as location and size are extracted and determined from the local disparities. Some experimental results are shown.

• 전기전자학회논문지
• /
• 제18권1호
• /
• pp.134-139
• /
• 2014

노인의 생활안전 사고 유형 중 가장 높은 비율을 차지하는 낙상은 57.2%이상이 가정에서 발생하는 것으로 조사되었다. 본 연구에서는 실내 바닥의 진동을 측정, 분석하여 낙상의 유무를 판별하고자 하였으며, 이를 위해 압전필름과 연산증폭기로 증폭 및 필터링 회로를 제작하여 진동 센서 모듈을 구성하였다. 진동 센서 모듈에서 증폭 및 필터링 과정을 거친 진동 신호는 데이터 수집 장치를 통해 디지털 신호로 변환되어 PC로 전송된다. 진동 신호는 k-NN 분류기를 이용하여 낙상 유무를 판별한다. 피험자 10명을 대상으로 낙상 실험결과, 분류기는 93.6%의 인식율을 나타내었다. 제작된 센서 모듈은 낙상 검출에 유용한 것으로 판단된다.

• 제어로봇시스템학회논문지
• /
• 제21권9호
• /
• pp.858-862
• /
• 2015

This paper presents a new forward-viewing mono-camera based obstacle detection algorithm for mobile robots. The proposed method extracts the coarse location of an obstacle in an image using inverse perspective mapping technique from sequential images. In the next step, graph-cut based image labeling is conducted for estimating the exact obstacle boundary. The graph-cut based labeling algorithm labels the image pixels as either obstacle or floor as the final outcome. Experiments are performed to verify the obstacle detection performance of the developed algorithm in several examples, including a book, box, towel, and flower pot. The low illumination condition, low color contrast between floor and obstacle, and floor pattern cases are also tested.

• 한국지리정보학회지
• /
• 제6권3호
• /
• pp.95-106
• /
• 2003

본 연구에서는 최근 도시계획, 수자원 및 방재 분야에서 높은 관심을 가지고 연구중인 항공레이저측량기법을 활용하여 시가지의 건축밀도를 평가하였다. 먼저, 레이저스캐너인 LiDAR(light detection and ranging)로 취득한 표고자료로부터 DEM(digital elevation model)과 DSM(digital surface model)을 구축하였으며, 건물의 높이를 계산하기 위해 DEM에는 ZONALMEAN 필터, DSM에는 ZONALMAJORITY 필터를 적용하였다. 필터링 과정으로부터 계산한 층수와 현지조사의 층수를 비교한 결과 표준오차 ${\pm}0.199$층을 확보할 수 있었다. 또한 건물 연면적 레이어와 구획 레이어를 중첩하여 통계분석함으로서 도시 구획별 용적률을 제시할 수 있었다. 항공레이저측량 자료를 이용하여 계산한 용적률과 현지조사로부터 계산한 용적률의 비교 결과, 용적률의 표준오차를 ${\pm}2.68%$로 확보할 수 있었다. 따라서, 항공레이저측량 자료는 향후 토지이용계획을 수립하는 의사결정자에게 매우 유용한 자료를 제공해 줄 것으로 기대한다.

• Smart Structures and Systems
• /
• 제14권4호
• /
• pp.595-615
• /
• 2014

This study aims to investigate the relationship between structural damage and sensitivity indices using the Hilbert-Huang transform (HHT) method. Two damage detection indices are proposed: the ratio of bandwidth (RB), and the ratio of effective stiffness (RES). The nonlinear four bays multiple degree of freedom models with various predominant frequencies are constructed using the SAP2000 program. Adjusted PGA earthquake data (Japan 311, Chi-Chi 921) are used as the excitations. Next the damage detection indices obtained using the HHT and the fast Fourier transform (FFT) methods are evaluated based on the acceleration responses of the structures to earthquakes. Simulation results indicate that, the column of the 1 st floor is the first yielding position and the RB value is changed when the RES<90% in all cases. Moreover, the RB value of the 1 st floor changes more sensitive than those from the top floor. In addition, when the structural response is nonlinear (i.e., RES<100%), the RB and the RES curves indicate the incremental change in the HHT spectra. However, the same phenomenon can be found from FFT spectra only when the stiffness reduction is large enough. Therefore, the RB estimated from the smoothed HHT spectra is an effective and sensitive index for detecting structural damage.

• 한국IT서비스학회지
• /
• 제10권1호
• /
• pp.105-116
• /
• 2011

With recent spotlight on the, uniquitous computing technology, the need for object of indentification and location infrastructure has increased. Such GPS technolgy must utilize IEEE 802.15.4 Zigbee used for existing wireless sensor network infra as a basice element for user's context-awareness in a uniquitous environement, for effectiveness.Such real-time GPS service is provided in the internal environment where the user would actually are and most high-rise buildlings apply. Underthe assumption, the real-time GPS technology is seperated by each floor, and signals do not get transmitted to other floors, the application on one floor within the high-rise buildling was conducted. This study intends to suggest a floor detection algorithm using IEE 802.15.3/Zigbee's RSSI which supports the accuracy within a couple of meters for the user's the movement between the floors in high-rise buildings in a complex environment. It proposes an floor detection algorithm using IEEE 802.15.4/Zigbee's RSSI which provides accuracy within a radius of few meters for the users movement between the floors for real-time location tracking within high-rise building in a cmoplex environment. Furthermore, for more accurate real-time location tracking, it suggests an algorithm for real-time location tracking using IEEE 802.15.4a/Zigbee's CSS technology based on triangulation. Based on the suggested algorithm, it designs a hybrid real-time location tracking service system in a high-rise buildling and test its functions.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.125-132
• /
• 1993

When a mobile robot moves around autonomously without man-made landmarks, it is essential to recognize the placement of surrounding objects especially for current position estimation, obstacle avoidance, or homing into the work station. In this paper, we propose a novel approach to recognize the floor paln for indoor mobile robot navigation using ultrasonic time-of-flight method. We model the floor plan as a collection of polygonal plane objects and recognize the floor plan by locating edges and vertices of the objects. The direction is estimated by the patterns of transmission beam and reception sensitivity of the ultrasonic transducer, and the distance is estimated by the correlation detection method. We show the validity of the proposed approach through experimental results and discuss the resolution and the accuracy of the estimation of direction and distance.

• 한국동물위생학회지
• /
• 제16권1호
• /
• pp.82-89
• /
• 1993

In order to detection of the intestinal parasites, fecal samples were taken from broiler (n=262), laying hen(n=244), parent stock(n=207) and native stock(n=287) in Chonbuk province. The prevalence and identification of intestinal parasites were determined by the fecal examination using the floatation and /or sedimentation methods and microscopical examination, respectively. The results were obtained as follows : 1. The detection rate of parasite-eggs from 4 flocks(total=1,000) was 65.7%. 2. In the breed and type of breeding, infection rate of parasite-eggs was detected 84.0％ as native stock (floor breeding, 241 chicken), 79.7％ as parent stock (floor breeding, 165 chickens), 73.3% as broiler(floor breeding, 192 chicken) and 24.2% as laying hen(cage breeding, 59 chicken), in order. 3. In the concern of mixed infection such as single, double and triple, the rate was 55.1%, 8.7% and 1.9%, respectively. 4. Ten kinds of infective eggs were isolated from 657 fecal sample of 4 flock. They were classified 51.l% as Eimeria spp., 12.7% as Ascaridia galli, 5.1% as Capillaria spp., 4.1% as Strongyloides avium, 2.3% as Heterakis gallinarum, 0.5% as Hymenolepis spp., 0.3% as Railleina spp. and 0.2% as Syngamus spp., Trichostrongilus spp., or Choanoteania spp., single or in combination.

• 국제학술발표논문집
• /
• The 4th International Conference on Construction Engineering and Project Management Organized by the University of New South Wales
• /
• pp.12-16
• /
• 2011

Building Information Modeling (BIM) is being widely spread in AEC industry worldwide and also in South Korea. Although the creation of digital model is better to be started at design stage, it can also improve the productivity of construction by simulating the actual construction process and environment. This paper presents application of BIM-based simulations related with design changes to transfer floors in 58-storey reinforced concrete office building. Transfer floor is not only a structurally important part of the building but also a challenging part of the actual construction in terms of sequence and period due to the complexity of the work. Preconstruction of rebar, mechanical, and plumbing is performed to review the construction drawings and to perform clash detection. Each item of application is evaluated for its effectiveness on actual construction and for the development potential.