Human-level Moving Object Recognition from Traffic Video

Fei Zhu1, 2, Quan Liu1, 2, Shan Zhong1 and Yang Yang3

  1. School of Computer Science and Technology, Soochow University Shizi Street No.1 Box 158
    Suzhou, China, 215006
    {zhufei, liuquan, 20134027005}@suda.edu.cn
  2. Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Jilin University
    Changchun, China, 130012
  3. Lund University
    Box 117, 221 00 Lund, Sweden
    yang.yang@med.lu.se

Abstract

Video preserves valuable raw information. Understanding these data and then recognizing objects and tagging them are crucial to intelligent planning and decision making. Deep learning provides us an effective way to understand big data with a human-level. As traffic video is characterized by crowded scene and low definition, it will be non-effective to deal with the whole image once. An alternative way is to separate image and determine a small window for each moving object. A Q-learning based moving object recognition approach, which firstly finds out moving object region and then uses a Q-learning based optimization method to determine the most compact region that contain the moving object, is proposed. The algorithms enable to detect the most compact rectangle around the moving object at near real-time speed. After that, a deep neural network is used to semantic tag the recognized objects. The experiment results show the algorithms work effectively.

Key words

Q-learning, deep learning, moving object recognition, traffic video, big data

Digital Object Identifier (DOI)

https://doi.org/10.2298/CSIS141114026Z

Publication information

Volume 12, Issue 2 (June 2015)
Year of Publication: 2015
ISSN: 1820-0214 (Print) 2406-1018 (Online)
Publisher: ComSIS Consortium

Full text

DownloadAvailable in PDF
Portable Document Format

How to cite

Zhu, F., Liu, Q., Zhong, S., Yang, Y.: Human-level Moving Object Recognition from Traffic Video. Computer Science and Information Systems, Vol. 12, No. 2, 787–799. (2015)