MATEC Web of Conferences
Volume 54, 20162016 7th International Conference on Mechanical, Industrial, and Manufacturing Technologies (MIMT 2016)
|Number of page(s)||7|
|Section||Image processing and visualization|
|Published online||22 April 2016|
A novel object detection technique for dynamic scene and static object
Shenzhen Key Laboratory of Information Science and Technology Graduate School at Shenzhen, Tsinghua University, Shenzhen, China
The object detection in video streams plays an important role in computer vision applications. The background subtraction, comparing each new frame to a model of the background, is one of the most popular method. However, the static background is practically impossible, and dynamic background makes the perfect object detection difficult. In addition, the problem of static object significantly affects the performance. In the paper, based on the visual background extraction (ViBe) algorithm, we presented a new method to deal with all problems. For the dynamic scene, we presented a new update mechanism to obtain the more robust model in dynamic regions. To address the static object issue, we cancelled the propagation mechanism in ViBe, designed an algorithm to detect region where it is always detected as foreground, and distinguished the static object from ghost with a self-designed measurement, which combined the knowledge of region contrast and edge. We described our method in full details, and compared it with other background subtraction techniques. Our experimental results show that the proposed method outperforms several state-of-the-art object detection approaches. In addition, it can process 60 frames per second on an Intel i5 3.1 GHz CPU with C++ code.
© Owned by the authors, published by EDP Sciences, 2016
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.