Small Target Detection and Recognition based on Template Matching

DOI

10.2991/icmmita-16.2016.178How to use a DOI?

Keywords

unit entropy; small target detection; template matching; background suppression

Abstract

A small target detection and recognition algorithm based on template matching is proposed to quickly detect and recognize small targets in a video whose size are from 10ž10 to 40ž40 pixels in complex background. Existing approaches cannot detect and recognize small target whose size is less than 40ž40 pixels in images taken by common cameras, because there is no obvious contour and texture in such small target images. The proposed algorithm combines image preprocessing, background subtraction, motion prediction, unit entropy and traditional template matching techniques to form a complete algorithm for tracking small targets. The proposed algorithm reduces the running time by narrowing the search space and reducing the number of matching points. Experimental results showed that the algorithm can accurately and fast recognize the small target with any angle rotation, while SIFT(scale-invariant feature transform) algorithm cannot calculate sufficient feature points to recognize such small targets.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tiansheng Zhao
PY  - 2017/01
DA  - 2017/01
TI  - Small Target Detection and Recognition based on Template Matching
BT  - Proceedings of the 2016 4th International Conference on Machinery, Materials and Information Technology Applications
PB  - Atlantis Press
SP  - 964
EP  - 968
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmmita-16.2016.178
DO  - 10.2991/icmmita-16.2016.178
ID  - Zhao2017/01
ER  -