Automatic Target Recognition Using Cross-Plot Technique
Author Information
Author(s): Wong Kelvin Kian Loong, Abbott Derek
Primary Institution: RMIT University
Hypothesis
Can Cross-plots of binary patterns serve as effective signatures for automatic target recognition?
Conclusion
The study concludes that Cross-plotting can efficiently produce a digital fingerprint of a target that correlates well with its identity in a target repository.
Supporting Evidence
- Cross-plotting can produce unique signatures for different binary patterns.
- The technique is robust against noise and distortion in images.
- Real-time identification of targets is achievable with low computational resources.
- Cross-plot signatures can effectively differentiate between similar targets.
- The method shows improved performance compared to traditional neural network approaches.
Takeaway
This study shows that a special method called Cross-plotting can help computers recognize targets in images quickly and accurately.
Methodology
The study developed a computational prototype for automatic target recognition using Cross-plot signatures, which involved preprocessing images, extracting features, and comparing signatures for identification.
Limitations
The approach may struggle with vague structural details of targets captured at greater distances, and the accuracy can be affected by poor image resolution.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website