InstanceMesh를 이용한 대용량 데이터 3D시각화 처리 기술
3D Visualization Technology for Mass-data using Instance Mesh
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
대규모 데이터 시각화는 클라우드컴퓨팅, 데이터센터의 증가 등 기술의 발전에 따라 시스템 모니터링, 사이버 보안 위협 탐지, 데이터센터 자원 관리 등 다양한 분야에서 핵심적인 역할을 하고있다. 하지만 방대한 양의 정보를 3D로 실시간 시각화 하는 데 성능 한계가 있다. 본 논문에서는이러한 문제를 해결하기 위해 인스턴스메시(InstanceMesh) 기술 기반의 3D 시각화 접근 방식을 제안한다. 제안된 기술은 단일 드로우 콜로 여러 데이터를 효율적으로 렌더링하여 CPU와 GPU 리소스를 효과적으로 활용한다. 실험 결과, InstanceMesh 활용 시 프레임 수 110fps와 1MB의 낮은 메모리 사용량을 당성하여 기존 기술 대비 대규모 데이터를 효과적으로 처리할 수 있음을 확인하였다. 이를 통해 향후 연구에서는 데이터 상호작용 기능을 강화해 실시간 분석과 예측 시뮬레이션등 다양한 분야에서 활용될 것으로 기대된다.
Abstract
Large-scale data visualization is playing a key role in various fields such as system monitoring, cybersecurity threat detection, and data center resource management due to technological advancements such as cloud computing and the increase of data centers. However, there is a performance limitation in real-time 3D visualization of a large amount of information. In this paper, we propose a 3D visualization approach based on InstanceMesh technology to solve this problem. The proposed technology efficiently renders multiple data with a single draw call, effectively utilizing CPU and GPU resources. Experimental results show that InstanceMesh can effectively process large-scale data compared to existing technologies, achieving a frame rate of 110 fps and a low memory usage of 1 MB. Through this, it is expected that future studies will strengthen the data interaction function and utilize it in various fields such as real-time analysis and predictive simulation.
참고문헌 (15)
- Park, Y., "Visualization-aware sampling for very large databases", 2016 IEEE 32nd International Conference on Data Engineering (ICDE), pp.755~766 (2016)
- Masood, Mirza Danial, "Benchmarking Performance of CPU and GPU through 3D Rendering", pp.2020
- Seok-Hyun Kim, "Rendering States Changing Costs Reducing Techniquefor Real-time 3D Graphics", Journal of the Korea Academia-Industrial cooperation Society, 10(8), pp.1843~1849 (2009)[DOI]
- Tae-Gwon Kim, "An Acceleration Technique of Terrain Rendering using GPU-based Chunk LOD", Journal of Korea Multimedia Society, 17(1), pp.69~76 (2014)[DOI]
- "WegGL Overview"[URL]
- "Three.js"[URL]
- "matplotlib"[URL]
- Jae-beom Park, "Design and implementation of the honeycomb structure visualization system for the effective security situational awareness of large-scale networks", Journal of the Korea Institute of Information Security & Cryptology, 24(6), pp.1197~1213 (2014)
- T. Taylor, "Visualization for Computer Security : Exploring NetFlow Data Using Activity Diagrams and Connection Bundles", 2009 Cybersecurity Applications & Technology Conference for Homeland Security, pp.39~44 (2009)
- Koutra, Danai, "Perseus: An Interactive Large-Scale Graph Mining and Visualization Tool", Proceedings of the VLDB Endowment, 8(12), pp.2824102~ (2015)[DOI]
- Useok Kwak, "The State of the Art in Visualizing Large Graph Data", Annual Conference of KIPS, 2017(4), pp.802~803 (2017)
- K. T. Claypool, "On frame rate and player performance in first person shooter games", Multimedia Systems, 13(1), pp.3~17 (2007)[DOI]
- Jeong, Ji Seong, "Design and Implementation of WebGL-Based Physics Education System Based on Virtual", Journal of the Korea Computer Game Society, 27(2), pp.41~48 (2014), Korea Computer Game Society[DOI]
- Lee, Ahyun, "Spatial Information Platform with VWorld for Improving User Experience in Limited Web Environment", Electronics, 8(12), pp.1411~ (2019)[DOI]
- Franke, Loraine, "Modern Scientific Visualizations on the Web", MDPI Informatics, 7(4), pp.37~ (2020)[DOI]