Recognition and tracking of target based on three-dimensional joint transform correlation
2008
Classical joint transform correlator (JTC) has been widely used in pattern recognition and target tracking. However,
when a real-time three-dimensional (3-D) object is captured as a two- dimensional (2-D) image, the key depth
information for discriminating different objects is lost. In some cases, two dissimilar objects may appear to be the same
appearance in their 2-D space. In this paper, a novel optical-electrical hybrid JTC is presented by analyzing a series of
2-D projections of the 3-D tested objects from different points of view. After encoding the key depth information into the
2-D power spectrum as the phase factor of the complex amplitude, we can obtain the 3-D joint power spectrum (JPS) and
the 3-D joint correlation output, respectively. The presence of the 3-D object and its precise spatial position can be
recognized by analyzing the correlation peak. Simultaneously, in order to reconstruct the original target images from the
finite discrete spectrum, the space parameters of the system are founded based on discussing the relation between the
number of the 2-D projections and the quality of the correlation output. Furthermore, the way to overcome the
shortcomings in the correlation output, such as a low correlation peak, a low peak-to-sidelobe ratio and a wide
correlation peak width, are discussed. The simulation result shows that the proposed method is effective for recognition
and tracking the targets distributed in 3-D space.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
1
Citations
NaN
KQI