• Journal of Korea Multimedia Society
• /
• v.25 no.6
• /
• pp.807-818
• /
• 2022

In order to further improve the accuracy and time efficiency of behavior recognition in intelligent monitoring scenarios, a human behavior recognition algorithm based on YOLO combined with LSTM and CNN is proposed. Using the real-time nature of YOLO target detection, firstly, the specific behavior in the surveillance video is detected in real time, and the depth feature extraction is performed after obtaining the target size, location and other information; Then, remove noise data from irrelevant areas in the image; Finally, combined with LSTM modeling and processing time series, the final behavior discrimination is made for the behavior action sequence in the surveillance video. Experiments in the MSR and KTH datasets show that the average recognition rate of each behavior reaches 98.42% and 96.6%, and the average recognition speed reaches 210ms and 220ms. The method in this paper has a good effect on the intelligence behavior recognition.

• International Journal of Internet, Broadcasting and Communication
• /
• v.13 no.2
• /
• pp.156-165
• /
• 2021

In this paper, we propose an histogram weighted centroid based Siamese region proposal network for object tracking. The original Siamese region proposal network uses two identical artificial neural networks which take two different images as the inputs and decide whether the same object exist in both input images based on a similarity measure. However, as the Siamese network is pre-trained offline, it experiences many difficulties in the adaptation to various online environments. Therefore, in this paper we propose to incorporate the histogram weighted centroid feature into the Siamese network method to enhance the accuracy of the object tracking. The proposed method uses both the histogram information and the weighted centroid location of the top 10 color regions to decide which of the proposed region should become the next predicted object region.

• Smart Structures and Systems
• /
• v.30 no.4
• /
• pp.371-383
• /
• 2022

A novel impact localization method is presented based on impact induced elastic waves in sensorized composite structure subjected to temperature fluctuations. In real practices, environmental and operational conditions influence the acquired signals and consequently make the feature (particularly Time of Arrival (TOA)) extraction process, complicated and troublesome. To overcome this complication, a robust TOA estimation method is proposed based on the times in which the absolute amplitude of the signal reaches to a specific amplitude value. The presented method requires prior knowledge about the normalized wave velocity in different directions of propagation. To this aim, a finite element model of the plate was built in ABAQUS/CAE. The impact location is then highlighted by calculating an error value at different points of the structure. The efficiency of the developed impact localization technique is experimentally evaluated by dropping steel balls with different energies on a carbon fiber composite plate with different temperatures. It is demonstrated that the developed technique is able to localize impacts with different energies even in the presence of noise and temperature fluctuations.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2021.05a
• /
• 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.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.399-401
• /
• 2022

POI (Point-of-Interest) 추천은 다양한 위치 기반 서비스에서 중요한 역할을 있다. 기존 연구에서는 사용자의 모바일 선호도를 모델링하기 위해 과거의 체크인의 공간-시간적 관계를 추출한다. 그러나 사용자 궤적에 숨겨진 개인 방문 경향을 반영할 수 있는 structured feature 는 잘 활용되지 않는다. 이 논문에서는 궤적 그래프를 결합한 시공간 인식 attention 네트워크를 제안한다. 개인의 선호도가 시간이 지남에 따라 변할 수 있다는 점을 고려하면 Dynamic GCN (Graph Convolution Network) 모듈은 POI 들의 공간적 상관관계를 동적으로 집계할 수 있다. LBSN (Location-Based Social Networks) 데이터 세트에서 검증된 새 모델은 기존 모델보다 약 9.0% 성능이 뛰어나다.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.671-672
• /
• 2015

Due to harsh conditions of disaster areas, understanding of current feature of collapsed buildings, terrain, and other infrastructures is critical issue for disaster managers. However, because of difficulties in acquiring the geographical information of the disaster site such as large disaster site and limited capability of rescue workers, comprehensive site investigation of current location of survivors buried under the remains of the building is not an easy task for disaster managers. To overcome these circumstances of disaster site, this study makes use of an unmanned aerial vehicle, commonly known as a drone to effectively acquire current image data from the large disaster areas. The framework of 3D model reconstruction of disaster site using aerial imagery acquired by drones was also presented. The proposed methodology is expected to assist rescue workers and disaster managers in achieving a rapid and accurate identification of survivors under the collapsed building.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.5
• /
• pp.259-265
• /
• 2023

Pedestrian detection is used to effectively detect pedestrians in various situations based on deep learning. Pedestrian detection has difficulty detecting pedestrians due to problems such as camera performance, pedestrian description, height, and occlusion. Even in the same pedestrian, performance in detecting them can differ according to the height of the pedestrian. The height of general pedestrians encompasses various scales, such as those of infants, adolescents, and adults, so when the model is applied to one group, the extraction of data becomes inaccurate. Therefore, this study proposed a pedestrian detection method that fine-tunes the pedestrian area by Refining Layer and Feature Concatenation to consider various heights of pedestrians. Through this, the score and location value for the pedestrian area were finely adjusted. Experiments on four types of test data demonstrate that the proposed model achieves 2-5% higher average precision (AP) compared to Faster R-CNN and DRPN.

• Archives of design research
• /
• v.18 no.3 s.61
• /
• pp.63-74
• /
• 2005

Modern society is diversified society and is under complicated situation as the boundary of each area has been disappeared. To understand and accept such complicated situation as widely as possible, it is required to understand interaction. of receiver with intertextual environmental arts as the structure of open text. This study examined interaction of environmental arts in terms of intertextual feature based on experience of receiver on combined element of different space and time, combination of genres. This is the concept of meaning personal experience or situation as receiver participates the process of completing art works, and set the fantasy, post-place and speciality of location and temporal-spatial expression method, as characteristics of intertextuality. Features of such experience elements are used as methodology of analyzing characteristics of each work. Feature of fantasy uses strategy of inducing spatial experience of receiver with dematerialization for post-place and expands the place where events occur with intervention of contingency and event situation. It suggests the spatial-temporal expression method as the features focusing on process and reflecting changes in spatial-temporal continuum and speciality of place emphasizing context of place. In conclusion, environmental arts needs to be deep rooted on complicated existence aspect of receiver beyond metaphysical dimension depending on presence and to accomplish conversion of awareness of supplying bisection of life from that place. By doing so, environmental arts can live textual life as it gets together with all other texts in terms of text dimension and creativity can be reborn as practical creativity in intertextuality rather than uniqueness. Such combination with other areas and acceptance of various aspects of receivers who see and experience this will result to creation of open works which can be create newly over and over again in multi-dimensional aspects.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.37 no.1
• /
• pp.12-22
• /
• 2019

This study derived the design aspects by carrying out the case study of Neunggok prehistoric remains park, Ansansingil historic park and Yongjuk historic park to which are taken measures to preserve undesignated cultural heritages after studying the related laws & regulations and the systems as the design conditions of historic park preserving buried cultural heritages. The results of the study are as follows. First, according to the laws & regulations related to the historical parks, the historic sites should be preserved and utilized at the same time and can have history-related facility spaces, squares, rest spaces, exercise spaces, education & culture space, and convenience spaces. Second, by the space organization and the circulation system emphasizing only the preservation of buried cultural heritages, the feature-preservation space and the functional space are separated and due to not accepting the usage behavior considering peripheral land use, the effectiveness of the historical park was low. Third, the passive feature-preservation methods such as the preservation of the exposed site in architectural methods, the reproduction of the dugout hut, and the planting Royal azaleas or displaying stone after covering up the location of the pit dwellings with soil and the usage mainly for viewing have weakened the identity of the historical park. Fourth, the fence preventing users' access interferes experiencing the features, and the vertical structure protecting the upper part of the exposed features has overwhelmed the landscape of the historical parks. Fifth, it was difficult to figure out the feature space only by the texts mainly on terminologies and the excavation photographs presented on the information signs which introduce the buried cultural heritages.

• Journal of Korean Neurosurgical Society
• /
• v.30 no.1
• /
• pp.47-53
• /
• 2001

Objective : This study was undertaken to evaluate operative results and prognosis according to preoperative clinical status and histopathological finding of spinal cord tumor. Methods : We analyzed of clinical feature, tumor location, histopathologic finding, operative results and prognosis in 55 patients with spinal cord tumor during last 10 years. Results : 1) The incidence of spinal cord tumors varies with the age of affected patients who are 2 to 75 years of age. Peak incidences were in the 5th & 7th decade of life, and the ratio of male to female was 1.2:1. 2) The most common histopathologic type was neurinoma(41.9%). 3) The tumors were located most frequently in the thoracic area(22 cases, 40.0%) and in the intradural extramedullary space(30 cases, 54.5%). 4) The most common initial clinical feature was pain in 20 cases(36.4%). For neurologic status on admission, 30 cases(54.5%) showed motor disturbance. 5) In radiologic studies, there were abnormal finding in 21 cases from plain X-rays among 37 cases. The entire 20 cases in when myelography was done showed subarachnoid blockade, either complete or incomplete. The magnetic resonance imaging, regard as the most accurate diagnostic method, revealed the exact location of the tumor and the relationship of the tumor with the adjacent anatomical structure. 6) The total removal was possible in 36 cases(65.5%), subtotal removal in 17 cases(30.9%) and biopsy in 2 cases (3.6%). Nineteen cases(90.5%) among 21 cases with preoperative radiculopathy group showed recovery or improvement, where as only 11 cases(36.7%) among 30 cases with preoperative motor weakness group showed recovery or improvement, with statistically significant difference(p<0.01). Pathologically, 26 cases(83.9%) among 31 cases of neurinoma and meningioma showed postoperative recovery or improved, but only 1 case(6.3%) among 16 cases of metastatic tumor, astrocytoma and ependymoma recovered. Postoperative complication noted in 5 cases(9.1%), and were noted postoperative hematoma, pneumonia, pulmonary edema and spinal cord infarction. Conclusion : Preoperative neurologic status and histopathologic finding are considered important factors of Postoperative outcome in patients with spinal cord tumor.