A trace acquisition method of small moving target

Abstract

This paper presents a new method for detecting small moving targets from consecutive image sequences. Image sequences are processed to detect targets, assuming that the data samples have been spatially registered. Firstly, Maximum Value Projection and normalization are utilized to reduce the data samples and reject the background clutter. Then, targets are detected using connected component analysis. At last, the velocities of the targets are estimated by centroid localization and least-square regression. A sliding neighborhood operation is performed before target detection. It can greatly reduce the computation while preserving as much target information as possible. The experimental results indicate that this method can efficiently detect the small moving targets and accurately track their traces. Both centroid locating precision and tracking accuracy are within a pixel.

10 Figures and Tables

Cite this paper

@article{Ma2010ATA, title={A trace acquisition method of small moving target}, author={Junyong Ma and Desheng Wen and Shaodong Yang}, journal={2010 3rd International Conference on Advanced Computer Theory and Engineering(ICACTE)}, year={2010}, volume={5}, pages={V5-450-V5-454} }