- 期刊
RGB-D Abandoned Object Detection Based on GrabCut Using Kinect
摘要
Intensity-based techniques for video surveillance (e.g., moving object detection and tracking, object classification, human motion analysis, and activity understanding) are vulnerable to noise and misleading results. Brightness changes and shadow are typically incorporated during the detection phase, resulting in problems. This paper presents a specifically-purposed and novel design for abandoned object detection (AOD) system using the low-cost Microsoft Kinect sensor. Our system encompasses: (1) fully automated abandoned object segmentation by introducing 3D GrabCut in surveillance cases, and (2) a robust AO detector by fusing RGB and depth information gathered by Kinect. We provide both quantitative and qualitative measurements that show our suggested method is effective.
延伸閱讀
- Chiu, Y. P. (2013). Super-resolution Reconstruction for Kinect 3D Data [master's thesis, National Chung Cheng University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0033-2110201613562833
- Chen, P. H. (2016). Object Model Reconstruction System with High-resolution Texture Mapping using RGB-D Sensor [master's thesis, National Chiao Tung University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0030-2212201711590322
- 黃博熙（2018）。Improve Alignment of RGB and Depth Images for KINECT V2〔碩士論文，逢甲大學〕。華藝線上圖書館。https://doi.org/10.6341/fcu.M0419125
- Santoso, P. S. (2016). Learning-based Pedestrian Detection Applied to RGB-D Images [master's thesis, National Chiao Tung University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0030-0803201714413901
- Lin, J. A. (2019). RGB-D Based Object Detection via Anchor Box with Multi-Reduced Region Proposal Network and Multi-Pooling [master's thesis, National Tsing Hua University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0016-0206202016184007